Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain

Transcription

1 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain A thesis submitted in partial fulfilment of the requirements for the Degree of Master of Medical Science in General Surgery at University of Otago, Dunedin New Zealand By Malsha Kularatna 2017

2

3 Abstract Abstract Background Right iliac fossa (RIF) pain is one of the most common reasons for acute presentation to general surgical services, and acute appendicitis is one of the most common underlying diagnoses. The clinical diagnosis of appendicitis continues to challenge clinicians and this is reflected in negative appendicectomy rates of up to 20%. Clinical predication rules (CPRs) are one method used to improve diagnostic accuracy of diagnose appendicitis and reduce negative appendicectomy rates. The APPEND score is a novel CPR that was developed at Middlemore Hospital. Aim The aims of this thesis were firstly to undertake a systematic literature review of published CPRs, and secondly to prospectively evaluate the performance of the APPEND CPR within a pathway dedicated to the management of RIF pain. Methods A systematic review of the published CPRs for the diagnosis of appendicitis in adults was undertaken. All studies that derived or validated a CPR were included and their performance was assessed using sensitivity, specificity and area under curve (AUC) values. The second part of the thesis was a comparative cohort study of the APPEND CPR within a clinical pathway dedicated to managing adults presenting with RIF pain. The primary endpoint was negative appendicectomy rate and the study was powered to detect a 7% difference. Secondary outcomes were length of hospital stay and number of radiological investigations (ultrasound and CT) performed. Staff satisfaction with the APPEND CPR was assessed by a survey. Results Thirty-four papers fulfilled inclusion criteria for the systematic review; 12 derived a CPR and 22 validated these CPRs. Analysis was limited by the heterogeneity and quality of included studies. The overall best performer in terms of sensitivity (92%), specificity (63%), and AUC values ( ) was the Acute Inflammatory Response (AIR) score but only three studies validated this CPR. Prospective evaluation of the APPEND CPR was performed on 437 consecutive adult patients presenting acutely with RIF pain to Middlemore Hospital over a 6-month period. The negative iii

4 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain appendicectomy rate in the prospective cohort was 9.2% (95% CI: 5.3%, 13.2%) compared to 19.8% (CI 16.2, 23.4%) in the retrospective cohort that did not use the APPEND CPR. After adjusting for multiple variables, the odds of a negative appendicectomy was 2.33 (95% CI; 1.26, 4.3, P value 0.007) in the retrospective cohort. An APPEND score of 5 was 87 % specific for ruling in appendicitis (PPV 94%) and a score of 1 was 100% sensitive in ruling out appendicitis (NPV 100%). There were more US scans but no significant difference in CT scans performed in the APPEND cohort. The length of stay was 0.9 days more in the APPEND cohort (p=< ). Survey respondents reported the APPEND CPR easy to use, but response rate was only 12%. Conclusion Twelve CPRs for the diagnosis of appendicitis in adults have been published. The AIR score appeared to perform best but further validation is required. In a comparative cohort study incorporating the APPEND CPR the rate of negative appendicectomy was reduced by more than 50%. Further validation of the APPEND CPR, including a comparison with the AIR, would be beneficial. iv

5 Preface Preface First and foremost, I would like to thank my husband Kasun Talwatta and my parents Nihal and Priyani Kularatna. They have been an endless source of encouragement and understanding. I feel truly fortunate knowing that they will never cease to support me unconditionally in all of my endeavours. This work would not have been possible without the guidance and mentorship of Professor John McCall and Mr Andrew D. MacCormick. I feel extraordinarily privileged to have them as my academic supervisors and mentors. I would like to sincerely thank the administrative staff at the South Auckland Clinical School and Middlemore hospital. It has been immensely memorable and enjoyable to work alongside such dedicated, professional, and wonderful individuals. v

6 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain vi

7 Table of Contents Table of Contents Abstract... iii Preface... v Table of Contents... vii List of Tables...xi List of Figures... xiii Chapter 1 Acute Appendicitis Introduction Epidemiology Anatomy Classifications used in appendicitis Aetiology Pathophysiology Clinical presentation and diagnostic strategies History and examination Laboratory investigations Radiological investigations Transabdominal ultrasonography Computerised tomography (CT) Magnetic resolution imaging (MRI) Diagnostic laparoscopy Management Non-operative management Operative management Clinical prediction rules Definition of a clinical prediction rule (CPR) CPRs for the diagnosis of appendicitis vii

8 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain 1.10 The development of the APPEND clinical prediction rule Retrospective validation of the APPEND clinical prediction rule Summary Aims of thesis Chapter 2 Clinical prediction rules for appendicitis in adults: Which is best? Introduction Methods Search strategy Inclusion and exclusion criteria Exclusion criteria for derivation studies Exclusion criteria for validation studies Selection of studies Data extraction and statistical analysis Derivation studies Validation studies Assessment of methodological quality of validation studies Results Study Selection Derivation studies Validation Studies Discussion Conclusion Chapter 3 Prospective Evaluation of the APPEND CPR within a RIF Pain Pathway Introduction: The APPEND Clinical Prediction Rule Rationale for Prospective Evaluation of APPEND CPR within a RIF Pain Pathway Aims Methods Study Design viii

9 Table of Contents Development of the RIF Pain Pathway with APPEND CPR Staff Satisfaction Survey Setting Participants Variables Data Sources and Measurement Prospective Cohort Ensuring Uniform Data Collection Sample Size Statistical Methods Results Participants Participant Demographics Comparison between Prospective and Retrospective Cohorts Primary Outcome Assessment: Negative Appendicectomy Rate Secondary outcome analysis Diagnostic indices obtained from prospective validation of the APPEND CPR Comparison of Length of stay Radiological investigations Results from online survey Discussion Limitations Strengths Future implications Conclusion Chapter 4 Summary References Appendix 1: The APPEND scoring system ix

10 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain Appendix 2: Staff satisfaction survey Appendix 3: Sticker given to admitting staff Appendix 4: Spreadsheet used to record findings x

11 List of Tables List of Tables Table 1.1 Classifications of appendicitis (23)... 3 Table 1.2 Causes of Appendiceal Lumen Obstruction Leading to Appendicitis... 4 Table 1.3 Differential diagnosis of acute appendicitis (24)... 6 Table 1.4 Signs and symptoms of appendicitis (57)... 7 Table 1.5 CPRs available for the diagnosis of suspected appendicitis in adults Table 1.6 Diagnostic indices of the APPEND CPR from retrospective validation Table 2.1 Search terms used Table 2.2 Quality assessment criteria for validation studies based on previously defined criteria by Wasson et al. (73) Table 2.3 Characteristics of studies and the clinical prediction rules from derivation studies Table 2.4 Variables incorporated within CPRs Table 2.5 Results for sensitivity, specificity and quality of external validation studies (ordered by score, cut off and quality score) Table 3.1 Information recorded for each patient Table 3.2 Condensed version of the spreadsheet used for data collection Table 3.3 Demographic data for prospective cases Table 3.4 Comparison of participant demographics between prospective and retrospective cohorts 52 Table 3.5 Likelihood of having a negative appendicectomy in the retrospective cohort compared to the prospective cohort Table 3.6 ROC statistics for APPEND CPR (a) and Alvarado score (b) Table 3.7 Absolute number of patients with and without appendicitis for each score Table 3.8 Diagnostic indices from prospective validation Table 3.9 Comparison of length of stay between prospective and retrospective cohorts Table 3.10 Multiple generalise linear analysis of factors contributing to the length of stay Table 3.11 Comparison of CT scan rates between prospective and retrospective cohorts Table 3.12 Number of respondents from each department Table 3.13 Responses to satisfaction survey xi

12 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain xii

13 List of Figures List of Figures Figure 1.1 Anatomical position (a) and variation in position of the appendix (1)... 2 Figure 1.2 Variations in anatomical position of the appendix (1)... 2 Figure 1.3 The APPEND clinical prediction rule Figure 1.4 ROC curve analysis for APPEND CPR from retrospective validation Figure 2.1 PRISMA flow diagram Figure 2.2 Forrest plot of sensitivity Figure 2.3 Scatter plot of sensitivity and specificity Figure 3.1 RIF pain pathway incorporating the APPEND CPR Figure 3.2 ROC analysis for APPEND vs Alvarado score xiii

14 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain xiv

15 Acute Appendicitis 1.1 Introduction Chapter 1 Acute Appendicitis Appendicitis is defined as inflammation of the vermiform appendix (23, 24) and is one of the most common causes of acute surgical illness. It has a variable presentation and significant overlap with other clinical conditions making it a challenging clinical diagnosis. Early diagnosis and treatment are important to reduce morbidity and mortality. On the other hand over diagnosis exposes patients to treatment related harms and expense from unnecessary procedures (25, 26). The high morbidity associated with missed or late diagnosis of appendicitis can result in overtreatment of patients, leading to high rates of negative appendicectomy (27-29). Clinical prediction rules are a potential tool that can be used to improve diagnostic accuracy and thereby minimise the burden associated with over- or under-diagnosis. 1.2 Epidemiology The approximate lifetime risk of appendicitis is 7% in females and 9% in males (23, 27, 30, 31). In the United States, 250,000 cases of appendicitis are reported annually, representing 1 million patient-days of admission (24, 27). The incidence of appendicitis is lower in underdeveloped countries, especially parts of Africa, and in lower socioeconomic groups (23, 27, 30). This has been attributed to dietary habits of these populations (23, 27, 32). Appendicitis was first described by Reginald Hubert Fitz in Since its first description, the incidence of appendicitis rose in western countries during the end of the 19th century with a subsequent decrease after the early part of the 20 th century (33, 34). Although the reasons for this increased incidence are uncertain, suggestions include changes in diet, improvement in hygiene and changing patterns of childhood infection (33, 34). While earlier publications indicate a peak incidence between the ages of ten to twenty years, recent publications have indicated that this mean age has increased over time to the second and third decades of life (23, 27, 31, 32, 35, 36). However, perforation is more common in infancy and in the elderly, when mortality rates are also at their highest. The overall mortality rate increases from 0.8 per 1000 in simple appendicitis to 5.1 per 1000 for perforated appendicitis (26). 1

16 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain 1.3 Anatomy The appendix is a narrow tubular true diverticulum which originates from the lower pole of the caecum (37). It is considered to be a vestigial organ (38). Figure 1.1 Anatomical position (a) and variation in position of the appendix (1) Figure 1.2 Variations in anatomical position of the appendix (1) The appendix has no fixed position. Figure 1.1 illustrates the anatomical position of the appendix within the right iliac fossa (RIF). Although a retrocaecal position is the most common anatomical position, (in approximately 64-70% of the population), variations are common (Figure 1.2). Thus, the course of the appendix, the position of its tip, and the difference in appendiceal position considerably changes clinical finding. 1.4 Classifications used in appendicitis Clinical classification of appendicitis is based on macroscopic examination of the appendix perioperatively and further microscopic examination on removal by a pathologist. The two major categories of classification are simple and complicated appendicitis. Complicated appendicitis can be broken down further into gangrenous and perforated with or without abscess formation. Table 1.1 highlights the microscopic and macroscopic findings of the appendix as it progresses through stages of inflammation from simple to complicated appendicitis. 2

17 Acute Appendicitis Table 1.1 Classifications of appendicitis (23) Classification Normal appedix Macroscopic changes - Normal pathology No visible changes - Acute intraluminal inflammation - Acute mucosal/ submucosal inflammation Simple appendicitis Complicated appendicitis - Gangrenous No visible changes No visible changes Congestion, colour changes, increased diameter, exudate, pus Friable appendix with purple, green, or black colour changes - Perforated Visible perforation - Abscess (pelvic/intrabdominal) Mass found during examination or abscess seen on preoperative imaging; or abscess found at surgery Microscopic changes Absence of any abnormality Luminal neutrophils with no mucosal abnormality Mucosal or submucosal neutrophils and/or ulceration Transmural inflammation, ulceration, or thrombosis, with or without extramural pus Transmural inflammation with necrosis Perforation; not always visible in microscope Transmural inflammation with pus with or without perforation Clinical relevance Consider other causes Might be the cause of symptoms, but consider other causes Might be the cause of symptoms, but consider other causes Likely cause of symptoms Impending perforation Increased risk of post-operative complications Increased risk of post-operative complications Adapted from: Bhangu A, Søreide K, Di Saverio S, Assarsson JH, Drake FT. Acute appendicitis: modern understanding of pathogenesis, diagnosis, and management. The Lancet.386(10000):

18 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain 1.5 Aetiology There are several different hypotheses regarding the aetiology of appendicitis. The most commonly accepted hypothesis is that of obstruction of the appendiceal lumen with secondary bacterial infection leading to appendicitis (24). Common causes of luminal obstruction are listed in Table 1.2. Table 1.2 Causes of Appendiceal Lumen Obstruction Leading to Appendicitis Faecoliths (39) Hyperplasia of appendiceal lymphoid follicles as a result of bacterial, viral, or parasitic infections (40) Crohn s disease (23) Altered composition of luminal mucous secondary to Cystic Fibrosis(41, 42) Carcinoid tumours (40) Foreign bodies (40) Blunt abdominal trauma (43) Psychological stress (44) Hereditary (45) Multiple other aetiologies have been proposed by opponents of the obstruction and bacterial invasion theory. These include, viral infection and low fibre diets. Viral infections are thought to cause mucosal ulceration and subsequent bacterial superinfection (40, 46-49). Low fibre diets slow intestinal transit time causing stool retention within the appendix which in turn increases susceptibility to infection and compromise of extramural appendiceal vascular supply due to the consequent distension (40, 46-49). It is unlikely that any single aetiological factor explains all causes. In fact, appendicitis is most likely caused by a combination of factors and disease progress is further influenced by individual variations in the host inflammatory and immune response. 1.6 Pathophysiology The obstruction and bacterial invasion theory states that the blind ending, narrow, tubular anatomical configuration of the appendix paired with its physiological function of mucous secretion causes rapid distension of the appendix once obstructed (23, 40). The resultant increased intraluminal pressure causes lymphovenous occlusion, precipitating an undrained reservoir for bacterial proliferation, further increasing the pressure within the appendicular lumen. This cycle of self-perpetuating distention, 4

19 Acute Appendicitis pressure, and obstruction of drainage then leads to transmural oedema of the appendix. The congestion and inflammation of transmural oedema constitutes simple appendicitis (Table 1.2). Periumbilical pain which migrates to the RIF is one of the most commonly described symptoms of appendicitis (28). This initial periumbilical pain in appendicitis is due to inflammation of the appendix, which like all visceral organs lacks somatic afferent pathways (23, 29, 47). Contraction of the appendix against an obstructed lumen or irritation of the appendix increases intraluminal pressure and activates afferent autonomic nerve fibres that enter the spine at the level of T10, causing vague pain referred to the periumbilical area. As the inflammation becomes trans-mural it reaches the parietal peritoneum, which is innervated by somatic nerves, and the pain becomes localised to where the appendix lies (50). The increasing dilation of the lumen and stretching of the appendiceal serosa is also thought to result in the associated nausea, vomiting and anorexia, which usually follows the onset abdominal pain (51). Terminal ileum and caecal irritation may also cause associated diarrhoea in some cases (23). Complications of appendicitis ensue as the distension within this closed loop system reaches pressures that impede arterial inflow leading to tissue ischaemia, infarction, and gangrene. The resultant loss of mucosal integrity allows bacterial invasion and translocation. Eventually breakdown of the appendiceal wall leads to perforation and spillage of its infected intraluminal contents leading to localised abscess formation or generalised peritonitis (52, 53). Anatomical variations and variations in individual immune responses means that pathophysiology of appendicitis can be variable. For instance, the inflamed retro-caecal appendix may be poorly localised with a lack of transition from epigastric to right lower quadrant tenderness. Pre- or post-ileal appendicitis may result in vomiting and diarrhoea being the predominant symptoms (23, 54). Paediatric and elderly patients who have limited physiological reserve and a poor immune system often presents with more non-specific symptoms. 1.7 Clinical presentation and diagnostic strategies Appendicitis remains a challenging clinical diagnosis. As explained in section 1.6, the anatomical variations in position and individual variations in immune responses results in inconsistency in the clinical presentation of appendicitis. The close proximity to other organs that produce similar symptoms when inflamed results in a number of differential diagnoses that can mimic appendicitis. Table 1.3 illustrates some common differential diagnoses for appendicitis. 5

20 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain Table 1.3 Differential diagnosis of acute appendicitis (24) Gastrointestinal Urological Gynaecological Other Mesenteric adenitis Right ureteric Ectopic pregnancy Pneumonia Intestinal obstruction colic Ruptured ovarian Pre-herpatic Gastroenteritis Right follicle pain on the Intussusception Terminal ileitis Acute cholecystitis pyelonephritis Urinary tract infection Torted ovarian cyst Salphingitis/pelvic inflammatory disease right 10th and 11th dorsal nerves Perforated peptic ulcer Meckel s diverticulitis Colonic/appendicular diverticulitis Pancreatitis Rectus sheath haematoma The mainstay of clinical diagnosis continues to be a thorough history, examination and basic laboratory markers. This may be further supplemented by radiological investigations in the form of an ultrasound scan(uss), computed tomography (CT) or magnetic resonance imaging (MRI). A definitive diagnosis, especially in females, may oftentimes require subjecting the patient to a diagnostic laparoscopy History and examination History and examination are key components in the clinic diagnosis of appendicitis. In most cases a thorough history and examination enables differentiation between acute appendicitis and the other differential diagnoses presented in Table 1.3. This in turn increases the pre-test probability of correctly identifying patients who require further investigations or surgery. Consequently, the number of 6

21 Acute Appendicitis patients subjected to the risks of radiation exposure and complications of surgery are minimised resulting in a reduction of the burden this disease places on our health care systems. Acute abdominal pain is the most common complaint in patients presenting with appendicitis (27, 28, 55). Generally, patients with simple acute appendicitis report pain over a 12- to 24-hour period (49). The most commonly described sequence of pain is initial vague colicky peri-umbilical pain which becomes constant and sharp as it migrates and localises to the RIF. This has been shown to occur in about 60% of cases (24). Examination usually reveals tenderness at RIF. The site of maximal tenderness is often said to be over McBurney s point, which lies two thirds of the way along a line drawn from the umbilicus to the anterior superior iliac spine (51, 53). Signs of localised peritoneal irritation include; involuntary guarding, percussion tenderness, rebound tenderness in the RIF and in some cases, a positive Rovsing s sign (Table 1.4) (51). Table 1.4 outlines some of the other common signs which can be elicited secondary local peritoneal irritation. Wagener et al demonstrated that a history of periumbilical pain that migrates and the presence of right lower quadrant pain has a high positive predictive value while the absence of right lower quadrant pain and similar previous pain in the history had a high negative predictive value (56). However, they also highlighted that there was no single finding that could completely exclude a diagnosis of appendicitis (56). Table 1.4 Signs and symptoms of appendicitis (57) Symptoms Signs Signs to elicit Peri-umbilical pain Migration of pain to RIF Anorexia Nausea Diarrhoea Pyrexia ( C and over) Localised tenderness in RIF Muscle guarding Rebound tenderness Dunphy s sign pain in RIF on coughing Rovsing s sing Pain in RIF on palpation of left iliac fossa Psoas sign patient will lie with right hip flexed for pain relief Obturator sign pain on flexion and internal rotation of hip 7

22 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain There are a large number of symptoms and signs which can occur in association with abdominal pain in appendicitis. The most commonly described symptoms are anorexia, nausea, vomiting and diarrhoea (54, 56). Associated early signs include fever and elevation of serum inflammatory markers due to the acute phase response (23). As complicated appendicitis ensues with ischemia and perforation, localised findings are overtaken by signs of generalised peritoneal inflammation. The patient often has a rigid abdomen with percussion tenderness and guarding across all four quadrants. If formation of a localised abscess has occurred this may result in the ability to palpate a localised mass in the RIF. The patient also shows signs of a systemic inflammatory response with an elevated temperature, heart rate, respiratory rate and drop in blood pressure. As described previously variations in the anatomy results in inconsistent presentations of appendicitis. Patients with a pre- or post-ileal appendix may present with nausea and vomiting as their predominant symptoms. A retrocaecal appendix may produce flank pain with a positive psoas stretch sign (24). If the tip of the appendix is oriented toward the pelvis urinary or defecation urgency may be the principal symptoms with then tenderness elicited on intra-abdominal palpation through a pelvic exam (in women) or a rectal exam (in men) (24, 56). Patients at the extremes of age may also present variable signs and symptoms. Paediatric patients who lack maturity of both their immune system and the ability to communicate may often seem withdrawn and exhibit a reduction in their feeding. Elderly patients may present with delirium or confusion. Consequently, delays in diagnosis due to the non-specificity of the presenting symptoms and signs at the extremes of age result in a higher incidence of complicated appendicitis (55, 58) Laboratory investigations Biomarkers of inflammation are used to supplement the findings from the history and examination. They are especially useful in equivocal cases but also in women, children and the elderly. Inflammatory markers such as the total white blood cell count, c-reactive protein (CRP) and neutrophils/granulocytes are most commonly used. In a meta-analysis by Andersson et al, white cells, granulocytes and CRP appeared to be the strongest discriminators with a receiver operator characteristic (ROC) areas of 0 78 to 0 75 (28). Similar results were reported in a meta-analysis by Yu et al with CRP showing the highest accuracy (59). Other biomarkers recently tested included bilirubin, 8

23 Acute Appendicitis pro-calcitonin and calprotectin. These have shown less promising results with a low sensitivity for detecting appendicitis (59-61). Along with these biomarkers, other tests are used to rule out alternative diagnoses presented in Table 1.3. These include β-human chorionic gonadotrophin (β-hcg) in females to exclude pregnancy, urinalysis to investigate the possibility of urinary tract infection or urolithiasis. Nevertheless, results from urinalysis need to be interpreted with caution, as up to 40% of patients with appendicitis may have an abnormality in the urine, such as microscopic haematuria that can results from peritoneal irritation adjacent to the bladder (24) Radiological investigations In equivocal cases of appendicitis, radiological investigations can aid in formulating the correct diagnosis. The most commonly used modalities include USS and CT, while MRI generally remains reserved for use in pregnant women Transabdominal ultrasonography Ultrasonography uses echoes of ultrasound pulses to delineate objects or areas of different density in the body. Its pooled diagnostic accuracy in appendicitis is moderate with a sensitivity of 86% (95% CI 83 to 88) and specificity of 81% (95% CI 78 to 84) reported in a recent meta-analysis (62). The major advantages of this modality is its relatively low expense compared to the other methods described below and the absence of exposure to ionising radiation. Among the disadvantages of ultrasound includes operator dependence or higher levels of accuracy, and variation in image quality in relation to body habitus. Its accuracy has been shown to be higher in children who have thinner musculature and less abdominal fat compared to adults (23). The need for a specialist operator also poses limitations with regards to access during afterhours and weekends in some institutions Computerised tomography (CT) CT scans combine a series of X-ray images taken from different angles to create cross-sectional images, or slices, of the bones, blood vessels and soft tissues. The use of CT for diagnosis of appendicitis has shown superior results when compared to ultrasound with a sensitivity of 94% (95% CI, 0.91 to 0.95) and specificity 95% (95% CI, 0.93 to 0.96) (62, 63). The routine use of CT for all 9

24 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain patients presenting with possible appendicitis has been associated with a negative appendicectomy rates as low as 6% in North America (64). The major disadvantages of routine CT use include exposure to ionising radiation, especially in children and young adults, and the relatively high cost per scan. CT is less operator dependent and provides a more detailed imaging of intra-abdominal organs compared to ultrasonography. Hence is it especially useful in older adults who have a higher risk of malignancy and right sided diverticulitis Magnetic resolution imaging (MRI) MRI provides high resolution images using magnetic fields. It has high levels of accuracy with sensitivity and specificity values of 97% (95% CI 92 to 99) and 99% (95% CI: 94% to 99%) respectively. Importantly, it has no risk of exposure to radiation (65). However, the high levels of expense associated with its use means that very few institutions worldwide can provide immediateaccess to MRI for investigating appendicitis. Thus, MRI is usually reserved for pregnant women to eliminate radiation exposure to the foetus Diagnostic laparoscopy The risks of perforation and associated high mortality of leaving an inflamed appendix in situ has often resulted in patients with a normal appendix undergoing surgery based on clinical suspicion. In unclear cases a diagnostic laparoscopy is usually performed to visualise pelvic organs as well as the appendix prior to its removal (37). This is especially helpful in females who may have a gynaecological cause for their presentation. A non-productive laparoscopy is a term used for a diagnostic laparoscopy which shows other pathological findings but no surgically treatable cause for presentation. 1.8 Management Management of appendicitis can be divided into two major categories, operative and non- operative management Non-operative management A number of clinical trials have investigated non-operative management of acute appendicitis with the use of antibiotics. The major benefit of non-operative management is the reduction in morbidity and 10

25 Acute Appendicitis costs associated with surgery (25, 66, 67). There are also concerns regarding the rate of recurrence, the progression of simple into complex appendicitis and missing an underlying malignancy. A Cochrane meta-analysis of five randomised control trials that compared the use of antibiotics to appendicectomy in patients with suspected acute appendicitis showed that 73.4% (95% CI: ) patients treated with antibiotics were well within two weeks without major complications, compared to 97.4% (95% CI: ) of patients who underwent appendicectomy (68). The recurrence rate in the antibiotic only group was approximately 17%. However, the results from this study were inconclusive based on a 20% margin of non-inferiority (68). Further meta-analyses and randomised controlled trials done since this Cochrane review have shown recurrence rates ranging between 15%- 30% (23, 69, 70). Further concerns raised by this meta-analyses included the heterogeneity in measurements of complications, the lack of measurements for minor complications, inadequate sample size for important subgroup analyses, and variable length of follow up limiting an accurate assessment of recurrence rates (68). Therefore, further studies of higher quality with better follow up are required before there is conclusive evidence to support non-operative management for patients with suspected appendicitis Operative management Appendicectomy remains the gold standard treatment for acute appendicitis. This may be performed with an open (OA) or laparoscopic approach (LA). Although appendicectomy is considered to be a relatively minor operation, all operations are associated with some risk. Overall complication rates of % are reported, including wound infection ( %), pelvic abscess (up to 9.4%) and a mortality rate of approximately 0.5% have been reported (23, 71). Both procedures also require a general anaesthetic, with its own inherent risks. Negative appendicectomy is a term used to describe an operation done for suspected appendicitis in which the appendix is removed but found to be normal on histological evaluation. The reported rate of negative appendicectomy varies widely but is commonly around 20% (24, 26). The costs of preoperative care, the surgery itself along with the postoperative complication means that this continues to impose a high cost. 11

26 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain 1.9 Clinical prediction rules Definition of a clinical prediction rule (CPR) Clinical prediction rules (CPRs) are tools which are designed to assist medical decision making. They are derived from systematic clinical observations (72, 73). Wasson et al defined a clinical prediction rule as one which (73); Has three or more predictive variables obtained from the history, physical exam and simple diagnostic tests. Provides a probability of an outcome or suggested a diagnostic/therapeutic course of action. Is not a decision analysis, decision tree or practice guideline. By quantifying various components of the history, exam and basic laboratory investigations, CPRs aim to formally test, simplify and increase a clinician s accuracy of diagnostic and prognostic assessments for a given disease (74) CPRs for the diagnosis of appendicitis The clinical diagnosis of appendicitis is challenging due to variations in clinical presentation and a broad differential diagnoses. Patients often present out of hours and the initial assessment of patients is made by junior surgical staff who lack clinical experience (23, 75, 76). Thus, CPRs have been developed as a tool that can be used to aid the diagnosis and management of appendicitis (75). Key elements from the history, examination and simple laboratory investigations are combined to ascertain the probability of appendicitis. A standardised method of assessment to reduce variability, reduce the number of unnecessary radiological investigations and the number of negative appendicectomies (5, 77, 78). Thus, CPRs aim to reduce morbidity and minimise the economic burden on the heath system. The first CPR for appendicitis was derived by Alvarado et al in 1976 (79). Since that time a number of others have also been derived mainly for use in adults presenting with suspected appendicitis. Table 1.5 summarises current CPRs available. 12

27 Acute Appendicitis Table 1.5 CPRs available for the diagnosis of suspected appendicitis in adults Author, (reference) study dates. (CPR name) Number of variables in CPR Range of score Application of score Alvarado, (79) to = Discharge 4-6 = Observe 7-10 = Operation Andersson, (80) , (AIR 1 ) 7 0 to 12 >8 = High probability 5-8 = Indeterminate <5= Low probability Andersson, (81) Dec Aug (modified AIR score) = Discharge 6-9 = Observe = Theatre Christian, (82) to 5 >=4 = Appendicitis <4 = Not appendicitis Fenyo, (83) to 70 >12 = Laparotomy = Re-evaluate <-16 = Non-operative Eskenlinen, (84) to 67.6 >57 = Appendicectomy = Non defined <50 = No Appendicectomy Goh, (85) to 7 As per Alvarado 13

28 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain Author, (reference) study dates. (CPR name) Number of variables in CPR Range of score Application of score = Intermediate risk <-20 = Low risk Lindberg, (86) dates not specified to 66 >0 = High risk = Grey zone <-26 = Low risk Sammalkorpi, (87) to = High = Intermediate 0-10 = Low Tzanakis, (88) to 15 >8 = Appendicitis <8 = Not appendicitis Van den Broek, (89) to = Observe 1 AIR = Acute inflammatory response Despite the number of CPRs available, there are still questions over their clinical utility. A detailed systematic review of the currently published CPRs used in appendicitis is contained in Chapter 2 of this thesis. However, in summary the heterogeneity in populations used, derivation methodology, pragmatic utility, variable quality of validation, and variable diagnostic accuracy makes it difficult to identify a single CPR that is superior to others for use in adult patients with suspected appendicitis. Hence, in 2015 a novel CPR called the APPEND score was derived from retrospective analysis of all patients who presented to Middlemore Hospital with RIF pain over a 12-month period. The APPEND CPR was then validated on an independent, though retrospective, cohort of patients presenting during a subsequent time period. The derivation and retrospective validation of the APPEND CPR are 14

29 Acute Appendicitis described below in section 2.3 (this work was performed by Mikarae et al, and is the subject of another thesis) (90) The development of the APPEND clinical prediction rule The APPEND CPR was developed from a retrospective evaluation of 855 patients who presented to Middlemore Hospital, Counties Manukau Health with RIF pain between November 2010 and December Middlemore Hospital is a tertiary public hospital based in South Auckland, New Zealand. It serves a catchment of 534,750 patients (which is approximately 11% of the total New Zealand population) (91). Consecutive patients fifteen years or older who presented with acute RIF or RLQ pain of less than seven days duration to the General Surgery service were included. Patients with generalised peritonism and those who had a previous appendicectomy were excluded. Data pertaining to predictive factors identified as significant in previous CPRs were recorded. This included patient characteristics patient characteristics (age, gender, ethnicity), characteristic of pain (duration of pain, pain migrating to the RIF or migratory pain) and associated symptoms (presence of anorexia, nausea or vomiting, diarrhoea, dysuria, subjective fevers), operation notes, pathology, imaging and initial laboratory results were reviewed by a medical doctor. Examination findings were based on the admitting resident doctor s assessment and the vital signs on presentation. Patients were categorised into two groups by their final diagnoses as appendicitis and non-appendicitis. A diagnosis of appendicitis was made on the basis of the histopathology report. Patients who were discharged with no clear diagnosis were coded as non-appendicitis if they were not readmitted within 30 days with appendicitis. Data analysis was performed with IBM SPSS Statistics Software version 21.0 (92). Initially, univariate analyses by Chi-square tests was used to assess the association between categorical predictors evaluated and the final diagnosis of appendicitis. Subsequently multiple regression analysis was applied to factors identified by the univariate analysis. A p-value of less than 0.05 was considered to be statistically significant. The initial univariate analysis identified male gender, migratory pain, anorexia, local peritonism, white blood cells (WBC) >11, neutrophils >7.5, CRP 15, subjective fever and temperature 38 C to be 15

30 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain significant. While WBC > 11, subjective fever and temperature 38 C were significant on univariate analyses, the multiple regression analyses did not confirm this. The acronym APPEND was developed to aid its ease of use in the clinical setting. It stands for Anorexia, Pain (migratory), Peritonism (localised), Elevated C-reactive protein (CRP) of more than 15x10 6, Neutrophilia of more than 7.5x 10 6 and Dude (male gender). The maximum score for each category is one giving a total score out of six, further facilitating its usability. The patients are categorised into low, moderate or high risk groups for appendicitis based on this score. A pictorial representation of this scoring system is outlined below (Figure 1.3). APPEND SCORE Relative Anorexia = 1 point 1 Migratory Pain = 1 point 1 Localised Peritonism = 1 point 1 Elevated CRP >15 = 1 point 1 Clinical Management Score of 0-1 = Low suspicion of appendicitis. Discharge with general practitioner (GP) follow up Score of 2-4 = Observe overnight. Consider USS for patients under 50, CT for patient over 50. Score of 5-6 = High risk of appendicitis. Consider diagnostic laparoscopy. Elevate Neutrophils > 7.5 = 1 point 1 Dude (male gender) = 1 point 1 Total score out of 6 points 6 Figure 1.3 The APPEND clinical prediction rule Retrospective validation of the APPEND clinical prediction rule The APPEND CPR was then retrospectively validated using a data set from December 2011 to February Baseline characteristics of the derivation and validation groups were compared. The 16

31 Acute Appendicitis same univariate and binary logistic regression statistical analyses used for the derivation were performed on the validation data. The APPEND score compares favourably with the well-known Alvarado score. Figure 1.4 demonstrates the ROC curves for both the APPEND score and the Alvarado score. The APPEND score has a better area under the ROC curve (0.845) compared with the Alvarado score (0.792), though they are not statistically significantly different. Diagnostic indices for the APPEND score are presented in Table 1.6. Using a cut-off value of 5, the positive predictive value (PPV) of the APPEND CPR is 86%. This suggests that if all patients with a score of five or six were managed surgically, the rate of negative appendicectomy would be 14% which is a reduction 6%. Using a cut-off value of 1 (i.e. a score of 0 or 1), the negative predictive value (NPV) is 96%. This means that if all patients with a score of 0 or 1 were discharged home after alternative diagnoses are ruled out, 4% would have a diagnosis of appendicitis. Figure 1.4 ROC curve analysis for APPEND CPR from retrospective validation 17

32 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain Table 1.6 Diagnostic indices of the APPEND CPR from retrospective validation Diagnostic indices of the APPEND score at different cut-off values Sensitivity Specificity PPV NPV Efficiency Summary The APPEND CPR with a maximum score of six was developed to use in adult patients who present with right lower quadrant pain to optimise diagnosing and managing appendicitis. It stratifies patients based on their level of risk giving the clinician a clearer idea of how best to further investigate and manage the patient. Its performance was good in retrospective validation with a sensitivity of 97.8% and NPV 95.6% for a cut off value of one, and a specificity of 99.6% and PPV 95.5% for a cut off above five for high risk of appendicitis. It also demonstrated a 7% reduction in the negative appendicectomy rate. However, further prospective validation of this CPR is required to ascertain the reproducibility of these results Aims of thesis In general, there are three steps involved in the development and testing of a CPR. These include; creation of the clinical decision rule, testing or validating the CPR in retrospect and prospectively assessing the impact of the CPR on clinical behaviour (74). The first two steps of this process have been described above. The primary aim of this thesis was to prospectively evaluate the performance of the APPEND CPR within a pathway dedicated to RIF pain. The first part of this thesis is a systematic review of the currently published CPRs for the diagnosis of appendicitis in adults. The second part of the thesis reports the prospective validation of the APPEND 18

33 Acute Appendicitis CPR within a clinical pathway for managing adult patients presenting acutely to Middlemore Hospital with RIF pain. 19

34 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain 20

35 Clinical prediction rules for appendicitis in adults: Which is best? Chapter 2 Clinical prediction rules for appendicitis in adults: Which is best? 2.1 Introduction As mentioned in chapter one, CPRs are one of the tools which can be used to aid the diagnosis of appendicitis. They are derived from systematic clinical observations and aim to reduce uncertainty by standardising the collection and interpretation of clinical data (72, 73). They have been shown to provide a more objective method of assessment and standardisation of care for patients with suspected appendicitis, thereby reducing the number of unnecessary operations and patient exposure to radiation (93). Although a number of CPRs have been derived for diagnosing appendicitis it is unclear which of these performs most reliably. The aim of this systematic review was to identify all current CPRs for the diagnosis of appendicitis in adults and assess their performance. 2.2 Methods This study was completed in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement (94) Search strategy A comprehensive literature search was performed in MEDLINE, EMBASE, Pubmed and Cochrane Central Register of Controlled Trials databases from inception to February The search strategy is outlined in Table 2.1. Studies were restricted to English language and humans only. The reference lists of all included and relevant review articles were also searched to identify further potentially eligible manuscripts. 21

36 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain Table 2.1 Search terms used Search terms (appendicitis or appendicectom$ or appendectom$ or right iliac fossa pain or rif pain or right lower quadrant pain or rlq pain) AND (nomogram$ or algorithm$ or guideline$ or decision or checklist$ or score or scores or scoring or probabilit$ or protocol$ or pathway$ or rule or rules or predictive) Inclusion and exclusion criteria Only studies that derived or validated the impact of a CPR for use in adults (>=15) presenting with RLQ pain, RIF pain or abdominal pain suspicious of appendicitis were included. For the purposes of this study, a CPR was defined as one that (72-74); Had three or more predictive variables obtained from the history, physical exam and simple diagnostic tests. Provided a probability of an outcome or suggested a diagnostic/therapeutic course of action. Was not a decision analysis, decision tree or practice guideline. Both CPR derivation and validation studies were included. A derivation study was defined as a study that described the method of how a new CPR was developed and explained how it should be applied in a clinical setting. A validation study assessed performance of an existing CPR by ascertaining the sensitivity, specificity and/or AUC. If derivation studies included an internal validation component, the validation component was excluded from the validation study analysis due to a high risk of potential bias (72) Exclusion criteria for derivation studies When assessing articles which derived a CPR, studies that modified an existing scoring system in order to generate a new scoring system were included if the new parameters and cut off values were clearly defined. There was no restriction on study design. Scores which were derived for use solely in paediatric, elderly, pregnant, or single gender populations and those that did not assess the primary outcomes of appendicitis versus non-appendicitis and/or required the use of neural networks were excluded. 22

37 Clinical prediction rules for appendicitis in adults: Which is best? Exclusion criteria for validation studies Studies that validated CPRs in elderly populations, a single gender only or included patients younger than 14 years were excluded. Studies that looked at a subset of the scoring system or only patients that had imaging were also excluded. Three studies that did not state the age of the participants were also excluded. Studies that included patients younger than 14 years of age with a separate analysis for adults were included Selection of studies The initial search, title and abstract screen were performed independently by two reviewers. Any discrepancy between the two reviewers was discussed with the senior author. A total of 224 articles were identified as relevant and underwent full text review. 2.3 Data extraction and statistical analysis Derivation studies Data from studies describing derivation of a CPR were extracted using a standard proforma. Study characteristics, derivation methodology, scoring systems characteristics (e.g. use of weighting, positive versus negative scoring) and variables comprising the CPR were recorded for each study Validation studies Extracted data from CPR validation studies were also extracted using a standard proforma. These included study design, results obtained for sensitivity, specificity, PPV, NPV, likelihood ratios and Area under curves (AUCs) values from receiver operating curve (ROC) analysis. When more than two cut off values were evaluated for the prediction of high risk of having appendicitis, only the cut off recommended in the original derivation paper was used for analysis. When sensitivity and specificity were not calculated in the validation studies, these were calculated from the data available using a two by two table. Forrest plot confidence intervals (CI) were calculated using the variance method for all studies to minimise bias (95). 23

38 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain Assessment of methodological quality of validation studies The quality of included validation studies was assessed and scored using 15 pre-defined criteria by Wasson et al (Table 2.2) (73). These criteria were specifically designed to asses articles describing clinical prediction rules. Table 2.2 Quality assessment criteria for validation studies based on previously defined criteria by Wasson et al. (73) Data collected prospectively 1 Study site well described (place, department, number of centres (2/3 have to be present) 1 Rule derived/validated on all patients at risk (i.e. all patients with abdominal pain at risk of appendicitis 1 Study population well described (age, sex) 1 Outcomes studied well defined (appendicitis vs no appendicitis), confirmation of appendicitis based on histology. 1 Blinding of those assessing predictors to outcome and/or vice versa 1 Adequate follow up of outcomes (phone call or evaluation of readmission from clinical notes for at least 30 days post admission) Predictors evaluated defined well uniform definition of signs and symptoms of the scoring system and agreement regarding clinical assessment methods. Mathematical techniques used reported; o descriptive - 1 o logistic regression - 2 o cross validation 3 95% confidence interval reported 1 Adequate precision estimation/power calculation 1 24

39 Clinical prediction rules for appendicitis in adults: Which is best? Adequate reporting of results (report four out of six of; Sensitivity, specificity, negative predictive value, positive predictive value, Likelihood ratio or Area under the curve) 1 Reproducible methodology Results Study Selection The initial database search identified 7696 titles, and a further 56 identified through the manual search. Of these 4398 were potentially relevant after removal of duplicates and further screening. Following abstract review 257 papers met criteria for full text review. Of these, 12 papers describing derivation of CRPs and 22 describing validation were included. The PRISMA flow diagram is presented in Figure 2.1 (94). 25

40 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain Figure 2.1 PRISMA flow diagram Derivation studies Characteristics of CPRs derived for use in adults with suspected appendicitis demonstrated significant heterogeneity in both study population and methodology (Table 2.3). Among the discrepancies in methodology was the variation in statistical analyses. Three studies used univariate analysis while seven studies used multivariate analysis (Table 2.3) (79-81, 83, 86-89, 96). The most commonly incorporated variable was the white cell count, which appeared in all 12 studies (Table 2.4) (79-83, 85-89, 96). Temperature, rebound tenderness and migratory pain were the next most common across all studies (Table 2.4) (79-83, 85, 89). Studies that used multivariate analysis identified gender, elevated C-reactive protein, RIF pain, neutrophilia, vomiting and signs of peritonism (guarding, rigidity) as likely variables (80, 81, 86-89). Rectal tenderness, diarrhoea, and Rovsing s 26

41 Clinical prediction rules for appendicitis in adults: Which is best? sign were the least commonly used variables, and appeared only in CPRs that used univariate analysis (83, 85, 96). Table 2.3 Characteristics of studies and the clinical prediction rules from derivation studies Author, study dates, name of CPR Numb er of patien ts Study design Mean age (year s) Patient populati on Derivatio n techniqu e Numbe r of variabl es Ran ge of score Weighti ng Application of score Alvarado, 305 Retrospecti 25.3 Abdomin Univariat 8 0 to Based on 0-3 = (79) ve al pain e 10 diagnosti Discharge 76 suggestiv e of appendici statistical analysis c weight 4-6 = Observe tis 7-10 = Operation Andersson, 316 Retrospecti 25.9 Suspecte Multivari 7 0 to Based on >8 = High (80) ve d acute ate 12 regressio probability 93, 1997 (AIR 1 ) appendici tis logistic regression n coefficient 5-8 = Indeterminate <5 = Low probability Andersson, 432 Prospectiv 21 Presentin Multivari 13 0 to Based on 5 = (81) Dec e g to ED ate 14 regressio Discharge Aug 2005 (modified AIR score) with lower abdomina l pain < 5 days logistic regression n coefficient 6-9 = Observe = Theatre Christian, 58 Prospectiv 24 Suspecte Derivatio 5 0 to None >=4 = (82) cases, e case- d acute n not 5 Appendicitis control s control study appendici tis described <4 = Not appendicitis 27

42 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain Author, study dates, name of CPR Numb er of patien ts Study design Mean age (year s) Patient populati on Derivatio n techniqu e Numbe r of variabl es Ran ge of score Weighti ng Application of score Fenyo, (83) 259 Prospectiv Can RLQ 2 Univariat Based on >12 = 1982 e be pain e to 70 likelihoo appendicecto used statistical d ratio of my in adults but no analysis each variable = Reevaluate mean age given <-16 = Non operative Eskenlinen, 1333 Prospectiv 38 Acute Multivari Based on >57 = (84) e abdomina ate to regressio Appendicecto 84 l pain logistic 67.6 n co- my regression efficient = Non defined <50 = No Appendicecto my Goh, (85) 238 Prospectiv Can Suspecte Two 5 0 to As per As per 2006 e review of be d acute variables 7 Alvarad Alvarado retrospecti used appendici removed o score ve data in tis from adults Modifie but no d mean Alvarado age score given Jahn, (96) 222 Prospectiv 27 Suspecte Univariat Based on >16 = High e study d acute e to 54 likelihoo risk 28

43 Clinical prediction rules for appendicitis in adults: Which is best? Author, study dates, name of CPR Numb er of patien ts Study design Mean age (year s) Patient populati on Derivatio n techniqu e Numbe r of variabl es Ran ge of score Weighti ng Application of score appendici tis from midnight statistical analysis d ratio of each variable = Intermediate risk <-20 = Low risk Lindberg, 746 Prospectiv Can Acute Multivari Based on >0 = High (86) dates e be abdomina ate to 66 likelihoo risk not specified used in adults l pain logistic regression d ratio of each variable = Grey zone but no mean age <-26 = Low risk given Sammalkor 829 Prospectiv Medi Suspecte Multivari 7 4 to Based on 16 = High pi, (87) 2014 e an 32 Rang e 25- d acute appendici tis or pain ate logistic regression 23 regressio n coefficie = Intermediate 47 in RLQ nt 0-10 = Low Tzanakis, 303 Prospectiv 28.3 Suspecte Multivari 4 0 to Based on >8 = (88) e d acute ate 15 regressio Appendicitis 2001 appendici tis logistic regression n coefficient <8 = Not appendicitis Van den 577 Prospectiv >11 Suspecte Multivari 5 0 to Based on 0-3 = Broek, (89) e d acute ate 9 regressio Observe appendici tis logistic regression n coefficient 4-6 = Diagnostic laparoscopy 29

44 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain Author, study dates, name of CPR Numb er of patien ts Study design Mean age (year s) Patient populati on Derivatio n techniqu e Numbe r of variabl es Ran ge of score Weighti ng Application of score 1 AIR = Acute inflammatory response 2 RLQ = Right Lower Quadrant 30

45 Clinical prediction rules for appendicitis in adults: Which is best? Table 2.4 Variables incorporated within CPRs Alvarado (79) Andersson (80) Andersson (81) Christian (82) Eskenlinen (84) Fenyo (83) Goh (85) Jahn (96) Lindberg (86) Sammalkorpi (87) Tzanakis (88) Van-denBoek (89) Total Gender 5 Right iliac fossa pain Other abdominal pain Onset of pain Migratory pain Duration of pain Intensity of pain Worse with movement Worse with cough Nausea 2 Vomiting 6 Anorexia 3 Diarrhoea 1 RIF tenderness Other abdo tenderness

46 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain Rebound tenderness Alvarado (79) Andersson (80) Andersson (81) Christian (82) Eskenlinen (84) Fenyo (83) Goh (85) Jahn (96) Lindberg (86) Sammalkorpi (87) Tzanakis (88) Van-denBoek (89) Total 8 Rigidity 4 Guarding 5 Rectal tenderness Rovsing s sign Temperatur e White cell count Neutrophils 4 C-reactive protein (CRP) 3 Others Left shift * Subjective fever Progression of pain * chemokine C C motif, ligand, chemokine ligand, interleukin, matrix metalloproteinase, myeloperoxidase, SAA serum amyloid 32

47 Clinical prediction rules for appendicitis in adults: Which is best? Validation Studies The 22 included validation studies demonstrated heterogeneity with respect to study population, study design and cut off values evaluated (Table 2.5). Two of the 22 studies only had AUC values available for the adult population. A scatter plot of all sensitivity and specificity values adjusted for sample size are shown in Figure 2.2. A Forrest plot could only be generated for sensitivity as the number of true negatives was unable to be calculated from the majority of the studies due to incomplete follow up of discharged patients (Figure 2.3). As CIs displayed in the Forrest plot were calculated using the variance method the values presented in Figure 2.3 may differ to those published in the original studies due to different calculation methods. The studies published by Scott et al and Erdem et al do not have CIs calculated as the sensitivity and sample size values were too similar. The majority of studies had a quality score between six and eight, while only six studies scored ten or more out of fifteen (Table 2.5). Of these, the two highest quality studies validated the acute inflammatory response (AIR) and Lintula scores (3, 15). A general trend demonstrated that at higher cut off values, the specificity of scoring systems improved but at the expense of the sensitivity. Clinically, this means CPRs with high cut off values are better for ruling in a diagnosis of appendicitis due to the good positive predictive value (Table 2.5, Figures 2.2 and 2.3). This was especially apparent in the Alvarado and AIR scores. The most commonly validated CPR was the Alvarado score, followed by the Kalan s modified Alvarado score (Figures 2.2 and 2.3) (1, 2, 4, 5, 8, 9, 11, 14, 16-19, 21, 22, 97-99). The average AUC value for the Alvarado score ranged between was higher than the modified Alvarado score which had an AUC of 0.69 from a single study (Table 2.5). The sensitivity of the Alvarado score ranged from 67.65% to 96.3% whilst specificity ranged from 58.18% to 89.39% when the originally recommended cut off of seven was used. This variability was also seen in Kalan s modified Alvarado score where the sensitivity ranged from 53.8% to 97.6%, and specificity ranged from 28.57% to 80% for the same cut off value. This variability remained regardless of the quality of the studies (Table 2.5). The AIR, Raja Isteri Pengiran Anak Saleha Appendicitis (RIPASA), Ohmman, Lintula and Eskelinen scores each had only a single validation study from which sensitivity and specificity could be obtained (3, 12, 15). 33

48 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain The AIR score showed a high sensitivity (92%) and moderate specificity (63%) at a cut off value above five. This reverted to 20% and 97% respectively for a cut off value above eight, which was the original cut off recommended by the authors. The AUC values generated for this CPR ranged from to 0.97, with an average value of (7, 10, 15). The Lintula score which was originally derived for use in paediatrics showed high performance in adults with a sensitivity of 87% and specificity of 96% (3). The final score looked at in this study was based on repeated calculations for patients who were observed as inpatients. This is in comparison to other studies which only reported diagnostic indices based on scores at admission. There were no AUC values available for this CPR. Erdem et al validated the Alvarado, RIPASA, Eskenlinen and Ohmann CPRs in a single study with a quality score of ten (12). While the RIPASA, Eskenlinen and Ohmann scores showed superior sensitivity and AUC values to the Alvarado scoring system, they showed poor specificity. The pragmatic utility of these scoring systems (table 2.5) demonstrated that the modified Alvarado score, Alvarado and AIR score are the most user friendly CPRs. The use of decimal points and multiple weightings make the other scores difficult to calculate in a busy clinical setting. 34

49 Clinical prediction rules for appendicitis in adults: Which is best? Table 2.5 Results for sensitivity, specificity and quality of external validation studies (ordered by score, cut off and quality score) Author Study design CPR validated Cut off value evaluated Number of patients in final analysis Tade (2) Prospective Alvarado Mean age ± standard deviation, median age + range, range 34 (17-56) Gender(% male) 63 Sensitivity Specificity 67.7 (79-91) AUC values Quality assessment score Owen (4) Prospective Alvarado Not given (92-100) Baidya/Rodrigues (14) Prospective Alvarado Gurav (16) Prospective Alvarado Limpawattanasiri (17) Pouget-Baudry (18) Prospective Alvarado Prospective Alvarado Pruekprasert (19) Prospective Alvarado Man (20) Prospective randomised Alvarado (16-65) Range 15-50, no mean given Range 15-72, no mean given 31.5 (range 15-88) Median 27 (14-75) 34.6 (17-87) (79-91) (73-87) (85-90) 92.8 (89-97) (74-84) Score group = (55-82) Yuksel (21) Prospective Alvarado ±10.8 (18-69) (38-50)

50 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain Author Study design CPR validated Cut off value evaluated Number of patients in final analysis Mean age ± standard deviation, median age + range, range Gender(% male) Sensitivity Specificity AUC values Quality assessment score Kim (1) Prospective Alvarado ±16.5; range 15 84) (36 49) 79.1 ( Kariman (5) Prospective Alvarado (No range given but participants were older than 16) (32-43) 98.4 (96-99) 10 Decastro (7) Prospective AIR 9 Alvarado (1-97 years), Separate AUC for adults Not available for adults =0.97, 50= 0.92, =0.88, 50 = Kollar (10) Erdem (12) Prospective Prospective AIR over Median = 0.805, 43 Not available for adults Alvarado (4-75) Alvarado RIPASA ± (range Ohmann Eskenlein > Scott (15) Prospective AIR 27 ± 16, 5 separate 92 (84-97) 63 (56-69) analysis for 9 ages (11-30) 97 (

51 Clinical prediction rules for appendicitis in adults: Which is best? Author Study design Lintula (3) Prospective randomised Meltzer (6) Prospective Battacharjee (8) Not stated Bulus (9) Prospective Malik (11) Prospective randomised casecontrol D Souza (13) Prospective Al- Hashemy (22) Prospective CPR validated Cut off value evaluated Number of patients in final analysis Lintula Modified Alvarado score >4 261 Mean age ± standard deviation, median age + range, range Score group 40 ± 17, nonscore group 41 ± (range 18-89) Gender(% male) Score group 50% non-score group 48% Sensitivity Specificity AUC values Quality assessment score 87 (81-94) (61-82) Modified Alvarado score Modified Alvarado score Separate analysis for adults Range 15-65, mean not available (71-88) (67-81) 39 6 Modified Alvarado Separate analysis for adults (74-90) Modified Alvarado score Modified Alvarado ± (16-76) (94-100) 53.8 (46-62)

52 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain Figure 2.2 Forrest plot of sensitivity Figure 2.3 Scatter plot of sensitivity and specificity 38

53 Clinical prediction rules for appendicitis in adults: Which is best? 2.5 Discussion There are currently 12 published CPRs available to aid diagnosis of adults presenting with suspected appendicitis. These have been validated in 22 separate studies. The aim of this systematic review was to ascertain which of these available scores performed the best. The heterogeneity of included studies precluded the possibility of performing a meta-analysis. Based on a narrative review, however, it appears the AIR score performs the best. Assessing the best performing CPR without meta-analysis meant narratively assessing sensitivity, specificity, AUC values, usability and the quality of available studies. Although the Lintula score performed highly in terms sensitivity and specificity, this score is difficult to use in a busy clinical setting and the comparability of the results obtained remains in question as the final score was based on repeated calculations as opposed to calculation at a single point in time. While the Eskenlinen, RIPASA and Ohmann scores had good sensitivity and AUC values they are difficult to calculate given the number of variables and range of weightings used. Thus the overall best performer in terms of the quality of studies, results and usability was the AIR score. It is easy to calculate manually and all parameters are easy to interpret except perhaps for the recommended subjective grading of rebound tenderness (as this requires clinical experience which may be limited in resident doctors). A score of five appears to be better than the originally recommended cut off of nine as there are a lower number of missed diagnoses without a significant reduction in specificity. The majority of published validation studies evaluated the Alvarado score and Kalan s modified Alvarado score. This is probably because Alvarado was among the pioneers to generate a CPR as a diagnostic aid for appendicitis (79). Although the Alvarado score is simple to calculate the interpretation of left shift in neutrophils is time consuming. The results from the available studies demonstrated wide variation for both sensitivity and specificity. This variation was further emphasised as cut off value increased and was also attributable to study design (e.g. prospective verses retrospective), variations in the characteristics of the evaluated patients, interpretation of variables of the CPR by different clinicians in different settings as well as the clinical expertise of the clinicians. While the overall sensitivity did not appear to show much variation between the Alvarado and modified Alvarado score, the specificity appeared to be lower for the modified Alvarado score (77, 79). Thus although the modified Alvarado score provides a more user friendly CPR, the removal of the left shift in neutrophils appeared to increase the number of false positives and was less accurate than the original CPR (77, 79). 39

54 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain Amongst derivation studies there was wide discrepancy in the derivation methodology used. Multivariate logistic regression is known to be more reliable than using univariate analysis (74). This is highlighted by those CPRs derived with the multivariate method consistently identifying variables used in clinical practice (76, 95, 100, 101). Variables such as rectal tenderness and diarrhoea that were identified in studies employing univariate analysis are seldom used clinically in the diagnosis of appendicitis (55, 76, 102, 103). The reliability of multivariate logistic regression analysis is further emphasised by CPRs which used this methodology such as the Lintula, AIR and Eskenlinen scores showing better sensitivity and AUC values compared to the Alvarado score which was derived using univariate analysis. Several studies investigating CPRs for appendicitis conclude that clinical judgement is comparable to CPR stratification, especially when performed by a senior surgeon (1, 19, 22, 98, 104). While this could imply that CPRs do not improve diagnostic accuracy compared to a senior surgeon, it provides evidence that CRPs can improve diagnostic accuracy to the level of an experienced surgeon when used by less experienced staff (13, 22, 98, 104). Given that resident doctors usually undertake initial evaluation of patients with suspected appendicitis, the use of a CPR is valuable in this context. Patient care is likely to be more standardised and unnecessary exposure to radiation and invasive investigations, including laparoscopy, minimised. The heterogeneity and quality of included studies precluded meta-analysis of available data. A further limitation was the predefined age criteria as many of the studies included children were excluded because the finding for children and adults could not be separated. The exclusion of non-english publications may also have excluded important validation studies done in other populations. 2.6 Conclusion There are currently 12 CPRs available for use in adults with suspicion of appendicitis. Heterogeneity in methodology and quality of available studies precluded a meta-analysis. The AIR score performed best in terms of sensitivity, specificity AUC values and usability but has been validated in only a small number of studies. The Alvarado and modified Alvarado were the most commonly validated CPRs but their performance was variable. The original Alvarado score outperformed the modified Alvarado score across all three criteria (sensitivity, specificity and AUC values). 40

55 Prospective Evaluation of the APPEND CPR within a RIF Pain Pathway Chapter 3 Prospective Evaluation of the APPEND CPR within a RIF Pain Pathway 3.1 Introduction: The APPEND Clinical Prediction Rule The APPEND CPR was developed and retrospectively validated on all patients who presented to Middlemore Hospital with RIF pain in 2015 by Mikarae et al (90). It incorporates key elements of the history, exam and laboratory investigations to provide a score out of 6 that predicts the likelihood of appendicitis. It aims to support the clinical diagnoses of acute appendicitis in patients who present with RIF pain (see Appendix 1 for a pictorial presentation). The retrospective validation showed a good performance overall, of this CPR, with a sensitivity and specificity of 97.8% and 99.6% (90). The NPV was 95.6% for a cut-off value of 1 (for low risk of appendicitis) and the PPV was 95.5% for a cut-off value of 5 (for high risk of appendicitis) (90). The final phase in development and testing of a CPR includes assessing the impact of the CPR on clinical behaviour (72, 74). Prospective validation allows assessment of a CPR s performance within a clinical setting. The prospective replication of the results obtained from a previous retrospective evaluation enhances the validity of a CPR (73, 74). The aim of this study was to prospectively validate the APPEND CPR within a clinical pathway. 3.2 Rationale for Prospective Evaluation of APPEND CPR within a RIF Pain Pathway RIF pain remains a very common reason for referral to general surgery and appendicitis is one of the most common causes, although there is a wide differential diagnosis (Table 1.3) (23, 105). As discussed in chapter one, appendicitis is predominantly a clinical diagnosis but has a variable presentation (24, 55, 106). This makes it a particularly challenging diagnosis for junior surgical registrars who lack the clinical experience and intuition that comes with seniority. Furthermore, the concern about the increased morbidity and even mortality associated with complicated appendicitis results in a high number of negative appendicectomies being performed (23, 25, 26). Negative appendicectomy is not only costly, it also carries some risk to patients (23, 71). The RIF pain pathway was created as a guideline for the investigation and management of the most common differential diagnoses of RIF pain. By creating an evidence-based, systematic and clear 41

56 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain clinical pathway that incorporated the APPEND CPR, we aimed to support the diagnosis of patients presenting with RIF pain. In doing so, we aimed to improve the outcomes for patients presenting with right iliac fossa pain at Middlemore hospital. 3.3 Aims The aim of this study was to ascertain whether the provision of clear guidelines and standardisation of care, including the routine use of APPEND CPR, would result in improved outcomes for patients presenting to Middlemore hospital with RIF pain. The primary outcome was the effect this pathway had on reducing the rate of negative appendicectomies at Middlemore Hospital. The secondary outcomes measured in this study were: The length of stay Use of radiological investigations Overall clinician satisfaction. 3.4 Methods The methods and results of this trial have been reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) criteria (107). The study was approved by the University of Auckland Human Participants Ethics Committee. Funding for this project was provided by a Faculty Development and Research Fund (FDRF) at the University of Auckland Study Design This was a comparative cohort study. Results from a prospective cohort managed according to the new RIF pain pathway were compared with the results obtained from the historical dataset collected for the retrospective validation of the APPEND CPR. 42

57 Prospective Evaluation of the APPEND CPR within a RIF Pain Pathway Development of the RIF Pain Pathway with APPEND CPR The APPEND CPR was incorporated into a clinical pathway. This was developed in conjunction with the Gynaecology, Emergency and Radiology departments. A simple and easy-to-follow flow diagram was created using data obtained from the retrospective study, stakeholder input as well as a systematic review of previous literature (Figure 3.1). The final pathway was discussed with the General Surgery and Gynaecology departments at Middlemore Hospital. Further discussions were held with Emergency and Radiology departments in order to ascertain the feasibility and areas requiring further improvement. 43

58 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain Patient referred to Gen Surg with RIF/RLQ pain 7 days, no previous appendicectomy, no generalised peritonism SOCRATES (Site, Onset, Character, Radiation, Associated symtpoms*, Timing, Alleviating/aggravating factors, Severity) APPEND SCORE 1. Anorexia (relative) 2. Pain (migratory) 3. Periotnism (localised) 4. Elevated CRP > Neutrophilia > Dude (males) Total score out of 6. Assoc symtpoms* i. Nausea ii. Vomiting iii. Diarrhoea iv. Distension v. Anorexia vi. Dysuria vii. Current time in cycle viii. Recent change in partner/unprotected sex ix. Change in nature of vaginal discharge x. Dysparaeuniae xi. Dyschezia xii. Dysmenorrhea A. Vitals pulse, BP, temp, RR, sats B. Tenderness vs local peritonism (guarding, percussion tenderness). C. Generalised peritonitis / Haemodynamic instability urgent discussion with senior for? OT or Further imaging Refer to gynae for further management + TVUS +/- MRI BHCG +/ known pregnancy Investigations a. MSU b. B-HCG (all females 16-50) c. FBC, U+E, LFT, CRP +/- G+H MSU ++ for RBC MSU ++ for WCC >100 or large numbers of bacteria Consider CTKUB for? Renal colic. Treat UTI B-HCG negative BhCG/MSU negative. Calculate APPEND score > 50 year old Score 2 or more 2 of x, xi or xii postive Gynae referrel for?endometriosis viii or ix Positive Gynae referrel for? PID - for examination + swabs Score 0-1 Discharge with appropriate return advise < 50 y.o. Score 2-4 Score 5-6 Observe ovenight + consider USS USS + Book for OT Diagnostic Lap +/- Appendicectomy Consider CT abdo for? Appendicitis vs. caecal diverticulitis vs maliganancy USS negative/ Score Recalculate APPEND score mane Score If temp 38 o C while waiting start cefuroxime + metronidazole Figure 3.1 RIF pain pathway incorporating the APPEND CPR 44

59 Prospective Evaluation of the APPEND CPR within a RIF Pain Pathway Staff Satisfaction Survey Clinical satisfaction was investigated using a survey that was created on Survey Monkey (Appendix 2) (108). This allowed members of all departments to provide feedback regarding the usability of the pathway and any further improvements required. The usability, understanding and ease of use were graded on a 4-point scale, where 1 was Extremely Easy and 4 was Not easy at all. A month after implementing the pathway, the staff satisfaction survey was sent to all members of the Surgical/Emergency and Gynaecology departments with regular reminders once a month. Staff were encouraged to respond once. The survey ran for six months, after which the results were tallied using cumulative charts and feedback identified Setting This study was conducted at the Department of General Surgery at Middlemore Hospital, which is a tertiary-level hospital in New Zealand serving a catchment of 534,750 patients (91). It was conducted from January to July Participants All patients aged fifteen and above referred to General Surgery at Middlemore Hospital were enrolled. The prospective cohort was managed as per the RIF pain pathway and APPEND CPR (Figure 3.1). This was compared to the same data recorded for the retrospective cohort. The retrospective cohort comprised of information for patients who presented with RIF pain over a one-year period from December 2011 to February Inclusion criteria for both cohorts were patients 15 years and older who presented with less than seven days of RIF pain. The exclusion criteria were patients with more than seven days of pain, patients who had an appendicectomy previously, or those with generalised peritonitis. The clinical records of patients from both cohorts were analysed for three months after their presentation for any returns/missed appendicitis or complications. 45

60 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain Variables Variables analysed included demographic data, clinical findings, laboratory and radiological investigations (Table 3.1). The calculated APPEND score was also recorded and the Alvarado score was calculated based on the information collected. Table 3.1 Information recorded for each patient Demographic data Age, gender, ethnicity Key information from history Duration of pain (hours), migratory pain, nausea, vomiting, anorexia, dysuria, vaginal bleeding, vaginal discharge, dysmenorrhea, dyspareunia, recent change in partner or unprotected sex Examination findings RIF tenderness, localised peritonism, testicular/vaginal examination Vital signs Findings from laboratory investigations Findings from radiological investigations Operative information Other clinical information Scores calculated Heart rate, respiratory rate, temperature, systolic blood pressure, diastolic blood pressure, oxygen saturations WCC, neutrophils, CRP, left shift in neutrophils, B-HCG, MSU results Histological results Chest x-ray, abdominal x-ray, USS or CT findings Surgical procedure performed, intraoperative diagnosis Length of stay in days, readmission and complications within the last 30 days APPEND, Alvarado 46

61 Prospective Evaluation of the APPEND CPR within a RIF Pain Pathway Outcomes analysed include: Negative appendicectomy rate Length of stay (measured in days from admission to discharge) The number of CT and USS scans performed Data Sources and Measurement Prospective Cohort All admitting surgical doctors were given cards containing the flowchart for their lanyards along with template stickers on which they could record their findings (Appendix 3). The sticker was placed on the patient s admission notes and the APPEND score calculated by the admitting registrar. The recorded data was entered into a spreadsheet by the primary investigator along with the demographic data, presenting vitals, results from laboratory and radiology investigations and final histology results. The same spreadsheet was used to record the data from the retrospective cohort (Table 3.2). An extract of the full spreadsheet is available in appendix 4. Table 3.2 Condensed version of the spreadsheet used for data collection NHI Age Gender Ethnicity Admission date Discharge date Length of stay Admit to XR Admit to USS Admit to CT Imaging to OT Admit to OT OT to home Consultant Team Pain duration Migratory pain Anorexia Nausea Vomiting Dysuria Left shift Lymph CRP Platelets Bili B-HcG MSU findings HV swab findings Endo swab findings CXR findings AXR findings XR date USS USS findings TV USS findings USS date CT CT findings Colon Colon findings Alvarado score 47

62 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain Fevers Diarrhoea PV bleed PV d/c Dysmenorrhoea Initial APPEND Evol Append Pre-op abx POST OP ANTIBIOTICS DURATION Surgical procedure performed Surg date ASA Dyspareunia Recent partner change Current time in cycle Procedure code RLQ tenderness Procedure 1 Localised peritonism Procedure 2 Rovsing Procedure 3 Hernia Procedure 4 Cx of hernia Gynae intra-op Testes exam Intra-op findings Testes exam findings Intra-op Dx PV d/c Intra-op Dx 2 Tender adnexa Histology Cx excitation Histology 2 Adnexal mass Histology 3 Temp Grade of appendicitis Febrile Cx HR Grade of cx Systolic BP Cx 1 Diastolic BP Cx 2 RR Major vs minor Cx O2 sats Readmission WBC Final diagnosis Leucocytosis Secondary diagnosis Neut Comments Ensuring Uniform Data Collection To ensure that the prospective data was recorded uniformly, all admitting registrars were briefed about the terminology and specific signs and symptoms. To minimise investigator bias, the APPEND score was recorded on a separate spreadsheet a week prior to recording the histology results. No identifying data was recorded for patients. Any missing data was filled in by the admitting registrars themselves. If all the information required was not available within 24 hours of the patient s admission, the patient was discarded from the prospective trial. Patients with missing data from the retrospective cohort were also discarded. 48

63 Prospective Evaluation of the APPEND CPR within a RIF Pain Pathway To minimise selection bias, an intention-to-treat approach was used. Thus, in cases where the consulting surgeon overrode the decision based on the APPEND score, the patient was still included in the final analysis Sample Size The retrospective sample contained 922 patients who presented with RIF pain. The prospective sample size calculation was performed using SAS version 9.4 (92, 109). Assuming that on average 922 patients present with RIF pain during the course of a year, with an average negative appendicectomy rate of 19.8% measured over a twelve-month period, before implementation, 384 patients were required in order to detect a 7% difference in negative appendectomy rates between pre- and post-implementation, with 85% power and 5% significance Statistical Methods Patients characteristics were represented in terms of mean with standard deviation (SD) for normally distributed continuous data. If the data was not normally distributed, median with Interquartile range (IQR) was reported and categorical variables were presented as counts and proportions. Bivariate analyses were carried out using Chi-square or Fisher s Exact test to determine any significant association between the two cohorts and the categorical variables. Parametric (two-sample t-test) and non-parametric tests (Mann-Whitney U test) were used to compare continuous variables between the two cohorts. Logistic regression model was used to determine significant association between the binary outcome appendicectomy and the cohorts groups. The logistic regression model was adjusted for covariates including age, gender, ethnicity, APPEND score, heart rate, systolic blood pressure and temperature. The results from univariate and multivariate logistic regression models were presented as odd ratios with 95% confidence intervals and p-values. Length of stay was analysed using a generalized linear model to evaluate whether there was any significant difference between the prospective and retrospective cohorts. Since the length of stay was found to be not normally distributed, a Gamma distribution was fitted. Normality and skewness of deviance and chi-square residuals were checked for model diagnosis. The model was adjusted with the variables age, gender, ethnicities and APPEND score. Significance level of less than 5% was considered statistically significant. The results from the gamma regression model was reported as 49

64 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain relative ratio with 95% confidence interval and p-values. All the analyses were carried out using SAS version 9.4 and SPSS version 22.0 (92, 109). To check for the accuracy and reliability of the diagnostic tool, diagnostic accuracy (sensitivity, specificity, NPV and PPV) was carried out using simple logistic regression that included final diagnosis of appendicitis as the dependent and APPEND score as the independent variable. Receiver Operative Curve (ROC) was produced with area under the ROC and its 95% confidence interval. The difference in CT and USS scan rates between the two cohorts were calculated using a two-proportion z-test. 3.5 Results Participants Participant Demographics 449 patients were enrolled in the study over a six-month period. Of these, 12 patients were excluded due to missing data. Thus, final analysis was performed on 437 patients. The retrospective cohort included 922 patients, of whom 67 had missing data. This gave us a total of 855 patients for our retrospective cohort. The demographics of all included participants for the prospective cohort are outlined below (Table 3.3). Overall the participants were aged between 15 and 82 with a mean age of 33. There was more or less an equal distribution of males (53.8%) to females (46.2 %) in the prospective cohort. 50

65 Prospective Evaluation of the APPEND CPR within a RIF Pain Pathway Table 3.3 Demographic data for prospective cases Age (mean ± standard deviation ) 33 (15) Gender (number + percentage) - Male 235 (53.8%) - Female 202 (46.2%) Ethnicity (number + percentage) - NZ European 113 (25.8%) - Other European 63 (14.4%) - NZ Maori 66 (15.1%) - Pacific Island 97 (22.4%) - Asian 51 (11.6%) - Indian 38 (8.7%) - Other 9(2.1%) Comparison between Prospective and Retrospective Cohorts A comparison of the demographic data between prospective and retrospective cohorts is shown in Table 3.4. Comparisons were made between demographic data, key vital signs (temperature, blood pressure and heart rate), APPEND and ASA scores. 51

66 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain Table 3.4 Comparison of participant demographics between prospective and retrospective cohorts Retrospective cohort Prospective cohort p value Age 35 (17) 33 (16) 0.03 Gender (% male) 356 (41.6%) 235 (53.8%) < Ethnicity (%) 1. NZ European 301 (35.2%) 113 (25.8%) 2. Other European 63 (7.4%) 63 (14.4%) 3. NZ Maori 137 (16.0%) 66 (15.1%) 4. Pacific Island 219 (25.6%) 97 (22.4%) 5. Indian 83 (9.7%) 51 (11.6%) 6. Asian 41 (4.8%) 38 (8.7%) 7. Others 11 (1.3%) 9 (2.1%) < APPEND score 2.8 (1.6) 3.1 (1.7) Heart rate (HR) 82 (16) 81 (16) 0.11 Systolic blood pressure (SBP) 129 (19) 128 (19) Temperature 37 (0.8) 36 (0.8) < ASA Score * I 389 (70%) 147 (67.1%) II 144 (26%) 60 (27.4%) III 17 (3%) 12 (5.4%) IV 2 (1%) 0 (0%) * ASA = American society of anaesthesiologist score There was no significant difference between the age, heart rate, systolic blood pressure or ASA score of the two cohorts. The retrospective cohort had a significantly higher number of males compared to the prospective cohort. The ethnicity compositions are also significantly different between these cohorts, with more Asian, Indian and other Europeans in the prospective cohort compared to the retrospective cohort. The APPEND score was higher in the prospective cohort. 52

67 Prospective Evaluation of the APPEND CPR within a RIF Pain Pathway The ASA scores between the 2 cohorts were similar with majority of the surgical patients being an ASA I or ASA II Primary Outcome Assessment: Negative Appendicectomy Rate Nineteen of 206 (9.2%; 95% CI: 5.3%, 13.2%) in the prospective cohort underwent a negative appendicectomy. The negative appendicectomy rate in the retrospective population was 19.8% (95% CI 16.2, 23.4%). Since there is no overlap in the confidence intervals of the negative appendicectomy rates between the two cohorts indicating a statistically significant difference (P value = from chi-square test). After adjusting for variables including age, gender, ethnicity, APPEND score, HR, SBP and temperature, the odds ratio of a negative appendicectomy for the retrospective cohort compared to our prospective cohort was 2.33 (95% CI 1.26, 4.3, P value 0.007) (Table 3.5). When comparing the two surgical cohorts the APPEND score was significantly associated with the negative appendicectomy rates. For each increment of the APPEND Score the odds of a negative appendicectomy reduced by Age was also shown to be significantly associated with negative appendicectomy rates where the odds of having a negative appendicectomy reduced with increase in age. The reduction in the negative appendicectomy rate was seen despite the inclusion of seven patients that underwent appendicectomy despite having a score of one, due to the supervising consultant overriding the protocol based decision. We included these patients in our final analysis in order to eliminate selection bias. None of these seven patients had appendicitis on histology. 53

68 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain Table 3.5 Likelihood of having a negative appendicectomy in the retrospective cohort compared to the prospective cohort Odds ratio 95% confidence interval p value Retrospective versus prospective Covariates Age Gender (Female versus Male) Ethnicity * 2 versus versus versus versus versus versus APPEND score < Heart rate SBP Temperature * 1= NZ European, 2= Other European, 3= NZ Maori, 4 = Pacific Island, 5= Asian, 6= Indian. 7 = Other Secondary outcome analysis Diagnostic indices obtained from prospective validation of the APPEND CPR The area under the ROC was 0.88 (95% CI: 0.85, 0.91) for APPEND score; and 0.86 (95% confidence interval: 0.82, 0.89) for Alvarado score. The ROC curves and associated statistics are shown in figure 3.2 and table

69 Prospective Evaluation of the APPEND CPR within a RIF Pain Pathway APPEND CPR Alvarado CPR Figure 3.2 ROC analysis for APPEND vs Alvarado score Table 3.6 ROC statistics for APPEND CPR (a) and Alvarado score (b) a) APPEND score ROC statistics Mann-Whitney ROC Model Standard 95% Wald Area Error Confidence Limits Model b) Alvarado score ROC statistics Mann-Whitney ROC Model Standard 95% Wald Area Error Confidence Limits Model The diagnostic indices obtained for each cut off point are shown in tables 3.7 and 3.8. At a cut off of zero to one only 2 cases of appendicitis were missed. Thus, the NPV for score of 2 or more was 98%. At a cut off of five the PPV was 94% resulting in only 6 cases of appendicitis which would have been missed. Thus at a cut off of five there is a low index of suspicion for missed appendicitis with a good predictive value for true appendicitis. These results were also comparable to the results obtained in our retrospective validation shown in table 1.6 in chapter one. The scoring system was found to be most efficient at a score above of four. i.e., the highest percentage of correct results were given at a score of four. However, this is at cost to the PPV of the test where by the PPV at a score of 4 is only 81%. 55

70 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain Table 3.7 Absolute number of patients with and without appendicitis for each score Append score Appendicitis No Yes Total % % 3.45% % 8.33% % 35.90% % 61.63% % 84.06% % 91.89% Total Table 3.8 Diagnostic indices from prospective validation Sensitivity Specificity PPV NPV Efficiency Comparison of Length of stay The comparison for length of stay between the two cohorts showed that there was a significantly increased length stay in the prospective cohort compared to the retrospective cohort, with a mean difference of 0.3-days (table 3.9). Table 3.9 Comparison of length of stay between prospective and retrospective cohorts Mean (days) Standard deviation median p25-p75 Retrospective cohort Prospective cohort Further analysis (table 3.10) shows that the odds ratio for the mean length of stay was 0.9 days less for the retrospective cohort compared to the prospective cohort. This equates to a 10% shorter length of stay in our retrospective population. Patients in the prospective cohort with negative diagnosis, the 56

71 Prospective Evaluation of the APPEND CPR within a RIF Pain Pathway elderly and males had a longer length of stay. The length of stay was inversely related to increase in the APPEND score. Table 3.10 Multiple generalise linear analysis of factors contributing to the length of stay Age group Ratio of Length of stay (days) 95% Confidence interval Lower boundary Upper boundary p value > 45 Reference age group APPEND score Female < Final negative diagnosis vs. positive diagnosis < Retrospective vs. prospective < Radiological investigations The results of the use of radiological investigation between the two cohorts are shown in table There was no significant difference between the number of CT scans performed in the two populations, however there were significantly more USS scans performed in the prospective cohort. Table 3.11 Comparison of CT scan rates between prospective and retrospective cohorts Imaging Retrospective cohort Prospective cohort P-value USS 114 (13.3%) (CI; 11.21, 15.78) 121 (27.7%) (CI; 23.7, 32.07) < CT 212 (24.8%) (CI; 24.13, 32.54) 123 (28.1%) (CI; 22.02, 27.8)

72 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain Results from online survey A total of nineteen out of 157 members responded to the survey. This comprised of nine out of sixtythree consultants, seven out of sixty-four registrars and three out of thirty house surgeons (table 3.12). Table 3.12 Number of respondents from each department Specialty Consultants Registrars House surgeons General surgery 4/15 3/30 3/30 Gynaecology 4/23 2/14 ED 1/25 2/20 The majority of respondents reported that the pathway was very easy to use and very useful clinically. The responses are summarised in table Table 3.13 Responses to satisfaction survey Q4: How satisfied Q1: How easy it is to Q2: How easy it is to Q3: How useful is are you with the use? understand? this pathway? pathway? Extremely easy (10.5%) Extremely easy (21.1%) Extremely useful (10.5%) Extremely satisfied (15.8%) Very easy (67.9%) Very easy (68.4%) Very easy (47.4%) Very satisfied (63.4%) Slightly easy (21.1%) Not so easy (10.5%) Slightly useful (36.8%) Dissatisfied (21.1%) Not easy (10.5%) Not at all easy (0%) Not at all useful Extremely dissatisfied (5.3%) (0.0%) While the majority of respondents had no further suggestions regarding further improvements required some comments left have been outlined below; Excellent concept I would propose changing the flow diagram by rearranging the boxes. Gynaecology Consultant. 58

73 Prospective Evaluation of the APPEND CPR within a RIF Pain Pathway Accuracy / reproducibility variable - some junior members interpret hx/ findings variably General Surgery Consultant. 3.6 Discussion The aim of this study was to ascertain whether the RIF pain pathway and the routine use of the APPEND CPR, would result in a reduction of the negative appendicectomy rate at Middlemore hospital. The results from this study shows that there was a significant reduction in the negative appendicectomy rates at Middlemore Hospital, from 19.8% down to 9.2%. The diagnostic accuracy of the APPEND score in the prospective cohort was comparable to that of the retrospective cohort in terms of the AUC, sensitivity, specificity, PPV and NPV. The NPV value for a score below two was 98% while the PPV for a score above five was 94%. On the other hand, there was no significant difference in the number of CT scans done between the cohorts. There was a statistically significant increase in the length of stay and the number of ultrasound scans done compared to the retrospective cohort. The primary hypothesis that the RIF pain pathway incorporating the APPEND CPR would lead to a reduction in the number of negative appendectomies was proven. While there is a lack of literature looking at clinical guidelines for appendicitis in adults, studies in paediatric populations have shown similar results in reducing negative appendicectomy rates (110, 111). A review on clinical guidelines by Woolf et al showed that clinical pathways offer clear, evidence based recommendations for clinicians who are unclear about how to proceed, overturn the beliefs of surgeons accustomed to outdated practices, improve the consistency of care, and provide authoritative recommendations that reassure practitioners about the appropriateness of their treatment (112). Thus, the RIF pain pathway provided the necessary guidance for the admitting surgical staff who often struggle to differentiate between general surgical, gynaecological and urological problems which can mimic appendicitis. This in turn allowed development of standardised, evidence based clinical practice when managing patients with possible appendicitis. As mentioned previously, CPRs provide a simplified method of increasing a clinician s accuracy of diagnostic and prognostic assessments for a given disease (74). Once other differential diagnoses were excluded, calculation of the APPEND score provided a further method of standardisation when triaging patients with the suspicion of appendicitis. Patients were assessed without the usual variations and bias seen in clinical practice for appendicitis (55, 76). It prompted early discussion with the operating senior surgeon for those in the high risk category while providing reassurance and encouragement towards observation and revaluation for those with intermediate scores. These 59

74 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain findings are in keeping with the results from our systematic review of previous CPRs which reported that evaluation based on CPRs were comparable to the decision making skills of a senior, experienced surgeon (13, 20, 22, 113, 114). The diagnostic accuracy of the APPEND CPR showed good correlation with the results obtained from retrospective validation. The decrease in the NPV with a reciprocal increase in the PPV as the APPEND score increased is consistent with findings published with other CPRs (100, ). Thus at lower values it remains a good rule out test while at higher values it is a good rule in test for appendicitis (116, 117).The AUC value for the APPEND score was superior to the Alvarado CPR and it was also superior to majority of AUC values of other CPRs recorded in the systematic review in chapter two. However, a direct comparison could not be made to the more superiorly performing AIR score as we did not collect sufficient data to calculate the AIR score in the prospective cohort. Although it was hypothesised that the length of stay would decrease, the opposite was found. There was a mean increase in length of stay of 0.3 days in our prospective cohort. This may have been due to an increased likelihood of observation and further investigations for those patients with an intermediate score (of two to four). The guideline recommended that patients under the age of fifty could be considered for an USS scan. This service is not available after hours at Middlemore Hospital. Overall, it is likely that the combination of these factors would have resulted in patients waiting longer for further review and/or an USS. While the increase in number of USS was expected given this recommendation for patients with intermediate scores, there was no significant difference in the number of CT scans performed. Previous assessment of the impact of clinical guidelines for appendicitis have consistently reported a reduction in the number of CT scans performed (110, 111, 118) This discrepancy is potentially explained by the recommendation that all patients above fifty years with an intermediate score should have a CT scan, along with its easier accessibility after hours. Overall the number of responses to the survey was low. While majority of responders found the pathway and CPR to be very useful, it may be that people who did not find the CPR useful or easy to use did not response to the survey creating a potential source of bias Limitations One of the major limitations of the study was that the derivation and prospective validation were all conducted at the same hospital. Thus the similar results obtained between the cohorts may be a consequence of location bias where the APPEND CPR worked particularly well for the patients who 60

75 Prospective Evaluation of the APPEND CPR within a RIF Pain Pathway present to Middlemore Hospital. Prospective validation in a different location would be needed to confirm the generalisability of the utility of the CPR. Furthermore, as the trial was conducted in the surgical department its applicability within a community or emergency department setting is uncertain. Only those with a higher suspicion were referred to the surgical department. Use of the same CPR in community or emergency department settings may not yield the same outcome. Although the use of the CPR was encouraged within the emergency department as well, keeping track of all patients from the community presenting with RIF pain was beyond the scope of this study. The collection of data by a single investigator meant that blinding of results could not be performed resulting in potential investigator bias. Furthermore, it also meant that eligible patients presenting after hours were only enrolled into the study if they were identified correctly by the admitting junior surgeon. This may have resulted in the loss of some patients who were discharged straight from ED, resulting in a lower number of true negative patients. Ideally the APPEND CPR should be compared to the AIR and Lintula scores which appeared to perform better than the Alvarado in the systematic review (115). Unfortunately, the validation studies of this systematic review was only analysed during the early phase of the trial and the spreadsheet for data collection had already been pre-designed based on the retrospective validation. This meant that there was inadequate information to calculate the results for the AIR and Lintula CPRs which in turn precluded direct comparison to them Strengths Prospective evaluation of a clinical prediction rule is thought to be the highest level of evidence for the validity of a CPR (74, 116). This study allowed us to identify the true reality of this CPR within a clinical situation. This study included all patients presenting with RIF pain. The majority of CPRs which have been validated to date have only included patients with a high suspicion of appendicitis. Thus the inclusion of all patients with RIF pain broadens the applicability of this pathway and scoring system when triaging patients. The primary investigator for the prospective phase was different from the primary investigator involved in the derivation and retrospective evaluation phase. This reduces, although does not eliminate, the risk of investigator bias. 61

76 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain In accordance with the intent-to-treat principle, we included seven patients that underwent appendicectomy despite having a score of one, due to the supervising consultant overriding the protocol based decision. We included these patients in our final analysis in order to eliminate selection bias. None of these patients had appendicitis on histology. The results obtained were adjusted for all clinical variables including the presenting vital signs. This included variables that did not show a significant difference increasing the reliability of our results when comparing the two cohorts. 3.7 Future implications Given the positive results we have obtained in standardising care and management of patients presenting with appendicitis, a multicentre study would allow assessment of the generalisability of the pathway and the APPEND CPR in different clinical settings. This would also enable referral between services to occur more rapidly based on standardised assessment. Given the feedback received from the survey and observations made during data collection and analysis, the pathway will be changed to emphasise the importance of further re-assessment and recalculation of the score for patients admitted for observation where by patients with an increasing score should be considered for theatre. In the study by Lintula et al a similar trend was seen whereby patients with an increasing score were found to have an increased likelihood of appendicitis while those with a decreasing score were safer for discharge (3). Collection of the extra data required to calculate the AIR and Lintula scores would also be beneficial as it would allow direct comparison of the APPEND score to these two scoring systems. Furthermore, it would be interesting to analyse the economic benefits of this pathway by calculating the total loss secondary to the increase length of stay and increase in number of USS versus the total saved with the reduction in negative appendicectomy rate. Application of the CPR would also be made simpler by changing it from a paper based to digital platform. 3.8 Conclusion The RIF pain pathway incorporating the APPEND CPR was successful in improving the diagnostic accuracy and management of patients referred to general surgery with RIF pain. In this prospective cohort study we showed a significant reduction in the number of negative appendectomies 62

77 Prospective Evaluation of the APPEND CPR within a RIF Pain Pathway emphasising previously stated benefits of the importance of evidence based standardisation of care for patients presenting with RIF pain. While further improvements regarding the cut off age for radiological investigations and management of patients kept in for further observation is required, the overall results obtained have been encouraging towards future use of this pathway and scoring system. 63

78 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain 64

79 Summary Chapter 4 Summary In an era where an evidence based approach to medical care is expected, the management of patients presenting with the common condition of RIF pain still lacks standardisation and consistency. While appendicitis remains the most common cause of RIF pain, there is a broad differential diagnosis and this poses a challenge to practitioners, particularly junior grade doctors who are responsible for the initial assessment (23, 119). The lack of standardised care for this common presentation may result in prolonged hospital admissions, unnecessary investigations and operations (112, 120, 121). This not only subjects patients to poorer quality of care, it also imposes higher costs to the health care system. Clinical practice guidelines and CPRS are tools designed to improve diagnostic accuracy and reduce variation in clinical practice (23, 113, 122). The first part of this thesis aimed to better understand of the CPRs described in the literature for the diagnosis of acute appendicitis. The second part of this thesis aimed to prospectively evaluate a novel CPR for appendicitis, known as the APPEND score, that was developed at Middlemore Hospital and incorporated into a clinical management pathway for patients presenting with RIF pain. The systematic review reported in chapter 2 identified twelve CPRs described in the literature to aid the diagnosis of appendicitis in adults (79-89, 96). Most of these CPRs were derived in patients with a high suspicion of appendicitis, or in patients with histologically proven appendicitis, thus limiting their applicability in a broader clinical setting. From a pragmatic point of view, some are also not particularly user friendly in an acute setting as they included over ten variables, decimal points or both positive and negative scoring. The Alvarado score is the oldest and most commonly validated CPR. Although the AIR and Lintula score appear to perform better, fewer validation studies have been performed on these more recently described CPRs. Thus, further prospective studies are required to determine the utility of these CPRs in clinical practice. The APPEND CPR was derived by Mikarae et al in 2015 from an audit on all patients who presented to Middlemore Hospital with RIF pain over a one-year period (90). A total of 855 patients were audited and multiple logistic regression analysis identified anorexia, migratory pain, localised peritonism, elevated CRP, neutrophilia and male gender as the six most significant factors in predicting appendicitis. These factors were abbreviated into an easy to remember acronym known as APPEND score. A retrospective validation showed good diagnostic accuracy in terms of sensitivity, specificity, PPV, NPV and AUC. It was also estimated that the APPEND score could potentially reduce the negative appendicectomy rate from 19.8% to 14% at Middlemore Hospital. 65

80 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain The aim of the second part of this thesis was to prospectively validate the APPEND CPR within a pathway for the management of RIF pain (chapter 3). A comparative cohort study was performed including all patients presenting with RIF pain from January 2016 to July Patients were managed according to a clinical pathway that incorporated the APPEND CPR and the outcomes were compared to the retrospective data. It was hypothesised that simplification and standardisation of care using this pathway and the APPEND CPR would reduce the negative appendicectomy rate by 7%. We also hypothesised that there would be a reduction in the length of stay, and the number of CT scans performed. Furthermore, we surveyed clinician satisfaction with the APPEND CPR. In the prospective cohort of 437 patients the negative appendicectomy rate was 9.2%, down from 19.8% in the retrospective cohort (p = ). In logistic regression analysis the odds of a negative appendicectomy was 2.33 (95% CI 1.26, 4.3, P value 0.007) times higher in the retrospective cohort compared to the prospective cohort, after adjusting for age, gender, ethnicity, presenting vital signs, and the APPEND score. The diagnostic indices obtained from the prospective analysis were comparable to the retrospective analysis and those reported in the literature (Chapter 2). The APPEND CPR compared well with other published CPRs. Furthermore, these results were obtained in a prospective intention to treat analysis whereas the majority of other CPRs have only been validated in a retrospective setting. This provides a higher level of evidence for the use of the APPEND CPR. However, direct comparison to the best performing CPRs in the literature, the AIR and Lintula scores, was not undertaken because we did not collect all the data required to collate these scores. The length of stay in the prospective cohort was not reduced and in fact increased by a mean of 0.3 days (i.e. approximately 8 hours), which is not considered to be clinically significant. Also, the anticipated reduction in the number of CT scans was not observed, which may be related to the age recommendation for CT and easier after hours availability of CT, compared with USS. The number of USS scans performed showed an increase in the prospective cohort. This was likely to reflect the recommendation in the clinical pathway to consider further radiological investigations for patient with intermediate scores. Finally, respondents to the survey found the APPEND score to be simple to use and helpful in clinical practice, however the response rate to the survey was only 12%. Incorporation of the APPEND CPR into a clinical pathway for the management of patients presenting to Middlemore Hospital with RIF pain was associated with a reduction in negative appendicectomies by 10.6%, a small increase in length of stay (0.3 days) and no change in use of CT scan. It appears to help standardise the approach to a challenging clinical diagnosis. Further improvements in the clinical 66

81 Summary pathway are possible, such as recalculation of the APPEND score in patients undergoing observation, and providing an easily accessible electronic platform. Further prospective validation will be needed to determine whether the APPEND score also performs well in other clinical settings. 67

82 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain 68

83 References References 1. Kim K, Rhee JE, Lee CC, Kim KS, Shin JH, Kwak MJ, et al. Impact of helical computed tomography in clinically evident appendicitis. Emerg Med J. 2008;25(8): Tade AO. Evaluation of Alvarado score as an admission criterion in patients with suspected diagnosis of acute appendicitis. West Afr J Med. 2007;26(3): Lintula H, Kokki H, Pulkkinen J, Kettunen R, Grohn O, Eskelinen M. Diagnostic score in acute appendicitis. Validation of a diagnostic score (Lintula score) for adults with suspected appendicitis. Langenbecks Arch Surg. 2010;395(5): Owen TD, Williams H, Stiff G, Jenkinson LR, Rees BI. Evaluation of the Alvarado score in acute appendicitis. J R Soc Med. 1992;85(2): Kariman H, Shojaee M, Sabzghabaei A, Khatamian R, Derakhshanfar H, Hatamabadi H. Evaluation of the Alvarado score in acute abdominal pain. Ulusal Travma ve Acil Cerrahi Dergisi. 2014;20(2): Meltzer AC, Baumann BM, Chen EH, Shofer FS, Mills AM. Negative predictive value of a low modified alvarado score for adult ED patients with suspected appendicitis. Academic Emergency Medicine. 2012;19:S de Castro SM, Unlu C, Steller EP, van Wagensveld BA, Vrouenraets BC. Evaluation of the appendicitis inflammatory response score for patients with acute appendicitis.[erratum appears in World J Surg Sep;36(9):2271]. World Journal of Surgery. 2012;36(7): Bhattacharjee PK, Chowdhury T, Roy D. Prospective evaluation of modified Alvarado score for diagnosis of acute appendicitis. J Indian Med Assoc. 2002;100(5):310-1, Bulus H, Tas A, Morkavuk B, Koklu S, Soy D, Coskun A. Can the efficiency of modified Alvarado scoring system in the diagnosis acute appendicitis be increased with tenesmus? Wien Klin Wochenschr. 2013;125(1-2): Kollar D, McCartan DP, Bourke M, Cross KS, Dowdall J. Predicting acute appendicitis? A comparison of the Alvarado score, the Appendicitis Inflammatory Response Score and clinical assessment.[erratum appears in World J Surg Jan;39(1):112; PMID: ]. World Journal of Surgery. 2015;39(1): Malik AA, Wani NA. Continuing diagnostic challenge of acute appendicitis: evaluation through modified Alvarado score. Aust N Z J Surg. 1998;68(7): Erdem H, Cetinkunar S, Das K, Reyhan E, Deger C, Aziret M, et al. Alvarado, Eskelinen, Ohhmann and Raja Isteri Pengiran Anak Saleha Appendicitis scores for diagnosis of acute appendicitis. World J Gastroenterol. 2013;19(47): D'Souza C, Martis J, Vaidyanathan V. Diagnostic efficacy of modified alvarado score over graded compression ultrasonography. Nitte University Journal of Health Science. 2013;3(3): Rodrigues G, Rao A, Khan SA. Evaluation of Alvarado score in acute appendicitis: a prospective study. The Internet Journal of Surgery. 2006;9(1): Scott AJ, Mason SE, Arunakirinathan M, Reissis Y, Kinross JM, Smith JJ. Risk stratification by the Appendicitis Inflammatory Response score to guide decision-making in patients with suspected appendicitis. British Journal of Surgery. 2015;102(5): Gurav P, Hombalkar N, Dhandore P, Hamid M. Evaluation of Right Iliac Fossa Pain with Reference to Alvarado Score Can We Prevent Unnecessary Appendicectomies? : JKIMSU;

84 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain 17. Limpawattanasiri C. Alvarado score for the acute appendicitis in a provincial hospital. J Med Assoc Thai. 2011;94(4): Pouget-Baudry Y, Mucci S, Eyssartier E, Guesdon-Portes A, Lada P, Casa C, et al. The use of the Alvarado score in the management of right lower quadrant abdominal pain in the adult. J Visc Surg. 2010;147(2):e Pruekprasert P, Maipang T, Geater A, Apakupakul N, Ksuntigij P. Accuracy in diagnosis of acute appendicitis by comparing serum C-reactive protein measurements, Alvarado score and clinical impression of surgeons. J Med Assoc Thai. 2004;87(3): Man E, Varga A, Lazar GY. Comparison of alvarado-score and clinical judgement in the diagnosis of acute appendicitis - Prospective, randomized trial. European Surgical Research. 2014;52 (3-4): Yuksel Y, Dinc B, Yuksel D, Enver Dinc S, Mesci A. How reliable is the Alvarado score in acute appendicitis? Ulusal Travma ve Acil Cerrahi Dergisi. 2014;20(1): Al-Hashemy AM, Seleem MI. Appraisal of the modified Alvarado Score for acute appendicits in adults. Saudi medical journal. 2004;25(9): Bhangu A, Søreide K, Di Saverio S, Assarsson JH, Drake FT. Acute appendicitis: modern understanding of pathogenesis, diagnosis, and management. The Lancet.386(10000): Hulme DJS, J. Acute appendicitis. British Medical Journal 2007;333: Ansaloni L, Catena F, Coccolini F, Ercolani G, Gazzotti F, Pasqualini E, et al. Surgery versus conservative antibiotic treatment in acute appendicitis: a systematic review and metaanalysis of randomized controlled trials. Digestive surgery. 2011;28(3): Blomqvist PG, Andersson RE, Granath F, Lambe MP, Ekbom AR. Mortality after appendectomy in Sweden, Annals of surgery. 2001;233(4): ADDISS DG, SHAFFER N, Fowler BS, Tauxe RV. The epidemiology of appendicitis and appendectomy in the United States. American journal of epidemiology. 1990;132(5): Andersson R. Meta analysis of the clinical and laboratory diagnosis of appendicitis. British Journal of Surgery. 2004;91(1): Andersson RE, Hugander AP, Ghazi SH, Ravn H, Offenbartl SK, Nyström PO, et al. Diagnostic value of disease history, clinical presentation, and inflammatory parameters of appendicitis. World journal of surgery. 1999;23(2): Buckius MT, McGrath B, Monk J, Grim R, Bell T, Ahuja V. Changing epidemiology of acute appendicitis in the United States: study period J Surg Res. 2012;175(2): Ceresoli M, Zucchi A, Allievi N, Harbi A, Pisano M, Montori G, et al. Acute appendicitis: Epidemiology, treatment and outcomes- analysis of consecutive cases. World J Gastrointest Surg. 2016;8(10): Arnbjornsson E. [Epidemiology of appendicitis]. Lakartidningen. 1983;80(45): Barker D, Morris J. Acute appendicitis, bathrooms, and diet in Britain and Ireland. British medical journal (Clinical research ed). 1988;296(6627): Barker D, Morris JA, Simmonds SJ, Oliver R. Appendicitis epidemic following introduction of piped water to Anglesey. Journal of epidemiology and community health. 1988;42(2): Ashley DJ. Observations on the epidemiology of appendicitis. Gut. 1967;8(6): Jess P. Acute appendicitis: epidemiology, diagnostic accuracy, and complications. Scand J Gastroenterol. 1983;18(2):

85 References 37. Hardin DM. Acute appendicitis: review and update. American family physician. 1999;60: Netter FH. Atlas of Human Anatomy: Elsevier Health Sciences; Nitecki S, Karmeli R, Sarr MG. Appendiceal calculi and fecaliths as indications for appendectomy. Surg Gynecol Obstet. 1990;171(3): Carr NJ. The pathology of acute appendicitis. Annals of Diagnostic Pathology. 2000;4(1): Coughlin JP, Gauderer MW, Stern RC, Doershuk CF, Izant RJ, Jr., Zollinger RM, Jr. The spectrum of appendiceal disease in cystic fibrosis. J Pediatr Surg. 1990;25(8): Oestreich AE, Adelstein EH. Appendicitis as the presenting complaint in cystic fibrosis. J Pediatr Surg. 1982;17(2): Hennington MH, Tinsley EA, Jr., Proctor HJ, Baker CC. Acute appendicitis following blunt abdominal trauma. Incidence or coincidence? Ann Surg. 1991;214(1): Creed F. Life events and appendicectomy. Lancet. 1981;1(8235): Andersson N, Griffiths H, Murphy J, Roll J, Serenyi A, Swann I, et al. Is appendicitis familial? Br Med J. 1979;2(6192): Arnbjörnsson E. Acute appendicitis and dietary fiber. Archives of Surgery. 1983;118(7): Arnbjörnsson E, Bengmark S. Role of obstruction in the pathogenesis of acute appendicitis. The American journal of surgery. 1984;147(3): Lamps LW. Beyond acute inflammation: a review of appendicitis and infections of the appendix. Diagnostic Histopathology. 2008;14(2): Prystowsky JB, Pugh CM, Nagle AP. Appendicitis. Curr Probl Surg. 2005;42(10): Caty M, Yamout S Appendicitis. Pediatric Gastrointestinal and Liver Disease (Fourth Edition). Saint Louis: W.B. Saunders; p Dunn JCY. Chapter Appendicitis. Pediatric Surgery (Seventh Edition). Philadelphia: Mosby; p Leung AK, Sigalet DL. Acute abdominal pain in children. Am Fam Physician. 2003;67(11): Ein SH. Appendicitis. In: Ashcraft KW, Murphy JP, Sharp RJ, editors. Pediatric Surgery. Philadelphia: W.B. Saunders; p Jess P. Acute appendicitis: epidemiology, diagnostic accuracy, and complications. Scand J Gastroenterol. 1983;18(2): Wagner JM, McKinney WP, Carpenter JL. Does this patient have appendicitis? Jama. 1996;276(19): Wagner JM, McKinney W, Carpenter JL. Does this patient have appendicitis? JAMA. 1996;276(19): Collins D. Bailey & Love's Short Practice of Surgery. 25th Edn N. S. Williams, C. J. K. Bulstrode and P. R. O'Connell (eds) mm. Pp Illustrated Hodder Arnold: London. British Journal of Surgery. 2008;95(10): Sallin K, Rothrock S. Diagnosis of acute appendicitis: increasing accuracy, improving outcome, and decreasing liability. The Journal of the Florida Medical Association. 1996;84(9):

86 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain 59. Yu CW, Juan LI, Wu MH, Shen CJ, Wu JY, Lee CC. Systematic review and meta analysis of the diagnostic accuracy of procalcitonin, C reactive protein and white blood cell count for suspected acute appendicitis. British Journal of Surgery. 2013;100(3): Burcharth J, Pommergaard H, Rosenberg J, Gögenur I. Hyperbilirubinemia as a predictor for appendiceal perforation: a systematic review. Scandinavian Journal of Surgery. 2013;102(2): Thuijls G, Derikx JP, Prakken FJ, Huisman B, van Bijnen Ing AA, van Heurn EL, et al. A pilot study on potential new plasma markers for diagnosis of acute appendicitis. The American journal of emergency medicine. 2011;29(3): Terasawa T, Blackmore CC, Bent S, Kohlwes RJ. Systematic review: computed tomography and ultrasonography to detect acute appendicitis in adults and adolescents. Annals of internal medicine. 2004;141(7): van Randen A, Bipat S, Zwinderman AH, Ubbink DT, Stoker J, Boermeester MA. Acute Appendicitis: Meta-Analysis of Diagnostic Performance of CT and Graded Compression US Related to Prevalence of Disease 1. Radiology. 2008;249(1): Florence M, Flum DR, Jurkovich GJ, Lin P, Steele SR, Symons RG, et al. Negative appendectomy and imaging accuracy in the Washington state surgical care and outcomes assessment program. Annals of surgery. 2008;248(4): Barger RL, Nandalur KR. Diagnostic performance of magnetic resonance imaging in the detection of appendicitis in adults: a meta-analysis. Academic radiology. 2010;17(10): Mason RJ, Moazzez A, Sohn H, Katkhouda N. Meta-analysis of randomized trials comparing antibiotic therapy with appendectomy for acute uncomplicated (no abscess or phlegmon) appendicitis. Surgical infections. 2012;13(2): Varadhan KK, Neal KR, Lobo DN. Safety and efficacy of antibiotics compared with appendicectomy for treatment of uncomplicated acute appendicitis: meta-analysis of randomised controlled trials Wilms IM, de Hoog DE, de Visser DC, Janzing HM. Appendectomy versus antibiotic treatment for acute appendicitis. The Cochrane Library Varadhan KK, Neal KR, Lobo DN. Safety and efficacy of antibiotics compared with appendicectomy for treatment of uncomplicated acute appendicitis: meta-analysis of randomised controlled trials. Bmj. 2012;344:e Salminen P, Paajanen H, Rautio T, et al. Antibiotic therapy vs appendectomy for treatment of uncomplicated acute appendicitis: The appac randomized clinical trial. JAMA. 2015;313(23): Sauerland S, Jaschinski T, Neugebauer EAM. Laparoscopic versus open surgery for suspected appendicitis. Cochrane Database of Systematic Reviews. 2010(10). 72. Laupacis A, Sekar N, Stiell l G. Clinical prediction rules: A review and suggested modifications of methodological standards. JAMA. 1997;277(6): Wasson JH, Sox HC, Neff RK, Goldman L. Clinical Prediction Rules. New England Journal of Medicine. 1985;313(13): McGinn TG, Guyatt GH, Wyer PC, et al. Users&#39; guides to the medical literature: Xxii: how to use articles about clinical decision rules. JAMA. 2000;284(1): Andersson RE. The magic of an appendicitis score. World Journal of Surgery. 2015;39(1): Beasly SW. Can we improve diagnosis of acute appendicitis? Bmj. 2000;321(7266):

87 References 77. Kalan M, Talbot D, Cunliffe WJ, Rich AJ. Evaluation of the modified Alvarado score in the diagnosis of acute appendicitis: a prospective study. Ann R Coll Surg Engl. 1994;76(6): Wilasrusmee C, Thakkinstian A, Proprom N. Diagnostic scores for appendicitis: A systematic review of scores performances. Journal of Gastroenterology and Hepatology. 2012;27: Alvarado A. A practical score for the early diagnosis of acute appendicitis. Annals of Emergency Medicine. 1986;15(5): Andersson M, Andersson RE. The appendicitis inflammatory response score: A tool for the diagnosis of acute appendicitis that outperforms the Alvarado score. World Journal of Surgery. 2008;32(8): Andersson M, Ruber M, Ekerfelt C, Hallgren HB, Olaison G, Andersson RE. Can new inflammatory markers improve the diagnosis of acute appendicitis? World Journal of Surgery. 2014;38(11): Christian F, Christian GP. A simple scoring system to reduce the negative appendicectomy rate. Ann R Coll Surg Engl. 1992;74(4): Fenyo G. Routine use of a scoring system for decision-making in suspected acute appendicitis in adults. Acta Chir Scand. 1987;153(9): Eskelinen M, Ikonen J, Lipponen P. A computer-based diagnostic score to aid in diagnosis of acute appendicitis. A prospective study of 1333 patients with acute abdominal pain. Theoretical Surgery. 1992;7(2): Goh PL. A simplified appendicitis score in the diagnosis of acute appendicitis. Hong Kong Journal of Emergency Medicine. 2010;17(3): Lindberg G, Fenyö G. Algorithmic diagnosis of appendicitis using Bayes theorem and logistic regression. Bayesian statistics. 1988;3: Sammalkorpi HE, Mentula P, Leppaniemi A. A new adult appendicitis score improves diagnostic accuracy of acute appendicitis - a prospective study. BMC Gastroenterology. 2014;14 (1) (no pagination)(114). 88. Tzanakis NE, Efstathiou SP, Danulidis K, Rallis GE, Tsioulos DI, Chatzivasiliou A, et al. A new approach to accurate diagnosis of acute appendicitis. World Journal of Surgery. 2005;29(9):1151-6, discussion van den Broek WT, Bijnen BB, Rijbroek B, Gouma DJ. Scoring and diagnostic laparoscopy for suspected appendicitis. Eur J Surg. 2002;168(6): Mikaere H, Zeng I, Lauti M, Kularatna M, MacCormick AD. Derivation and validation of the APPEND score: an acute appendicitis clinical prediction rule. ANZ J Surg Population profile: Counties Manukau district health board; 2015 [Available from: IBM SPSS Statistics for Windows, Version Armonk, NY:: IBM Corp; Ohle R, O'Reilly F, O'Brien KK, Fahey T, Dimitrov BD. The Alvarado score for predicting acute appendicitis: a systematic review. BMC Med. 2011;9: Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Annals of internal medicine. 2009;151(4): Zhou X-H, Obuchowski NA, McClish DK. Statistical Analysis for Meta-analysis. Statistical Methods in Diagnostic Medicine: John Wiley & Sons, Inc.; p

88 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain 96. Jahn H, Mathiesen FK, Neckelmann K, Hovendal CP, Bellstrom T, Gottrup F. Comparison of clinical judgment and diagnostic ultrasonography in the diagnosis of acute appendicitis: experience with a score-aided diagnosis. Eur J Surg. 1997;163(6): Huang TH, Huang YC, Tu CW. Acute appendicitis or not: Facts and suggestions to reduce valueless surgery. Journal of Acute Medicine. 2013;3(4): Man E, Simonka Z, Varga A, Rarosi F, Lazar G. Impact of the Alvarado score on the diagnosis of acute appendicitis: comparing clinical judgment, Alvarado score, and a new modified score in suspected appendicitis: a prospective, randomized clinical trial. Surg Endosc. 2014;28(8): Meltzer AC, Baumann BM, Chen EH, Shofer FS, Mills AM. Poor sensitivity of a modified Alvarado score in adults with suspected appendicitis. Annals of Emergency Medicine. 2013;62(2): Zhou X-H, Obuchowski NA, McClish DK. Regression Analysis for Independent ROC Data. Statistical Methods in Diagnostic Medicine: John Wiley & Sons, Inc.; p Zhou X-H, Obuchowski NA, McClish DK. Analysis of Multiple Reader and/or Multiple Test Studies. Statistical Methods in Diagnostic Medicine: John Wiley & Sons, Inc.; p Abou Merhi B, Khalil M, Daoud N. Comparison of Alvarado score evaluation and clinical judgment in acute appendicitis. Med Arh. 2014;68(1): Kasimov RR, Mukhin AS. Current state of acute appendicitis diagnosis. Sovremennye Tehnologii v Medicine. 2013;5(4): Yegane R, Peyvandi H, Hajinasrollah E, Salehei N, Ahmadei M. Evaluation of the modified alvarado score in acute appendicitis among iranian patients. Acta Medica Iranica. 2008;46(6): McCartan DP, Fleming FJ, Grace PA. The management of right iliac fossa pain Is timing everything? The Surgeon. 2010;8(4): Petroianu A. Diagnosis of acute appendicitis. Int J Surg. 2012;10(3): Moher D, Schulz KF, Altman DG, Group C. The CONSORT statement: revised recommendations for improving the quality of reports of parallel-group randomised trials. Elsevier; Survey Monkey Palo Alto, California, USA [ 109. SAS (r) Proprietary Software In: Inc SI, editor. 9.4 ed. Cary, NC, USA.: SAS Institute Inc; Warner BW, Kulick RM, Stoops MM, Mehta S, Stephan M, Kotagal UR. An evidencedbased clinical pathway for acute appendicitis decreases hospital duration and cost. Journal of Pediatric Surgery. 1998;33(9): Almond SL, Roberts M, Joesbury V, Mon S, Smith J, Ledwidge N, et al. It is not what you do, it is the way that you do it: impact of a care pathway for appendicitis. Journal of Pediatric Surgery. 2008;43(2): Woolf SH, Grol R, Hutchinson A, Eccles M, Grimshaw J. Potential benefits, limitations, and harms of clinical guidelines. British Medical Journal. 1999;318(7182): Zyluk A, Ostrowski P. An analysis of factors influencing accuracy of the diagnosis of acute appendicitis. Polski Przeglad Chirurgiczny. 2011;83(3):

89 References 114. Atema JJ, van Rossem CC, Leeuwenburgh MM, Stoker J, Boermeester MA. Scoring system to distinguish uncomplicated from complicated acute appendicitis. British Journal of Surgery. 2015;102(8): Kularatna M, Lauti M, Haran C, MacFater W, Sheikh L, Huang Y, et al. Clinical Prediction Rules for Appendicitis in Adults: Which Is Best? World J Surg Zhou X-H, Obuchowski NA, McClish DK. Measures of Diagnostic Accuracy. Statistical Methods in Diagnostic Medicine: John Wiley & Sons, Inc.; p Zhou X-H, Obuchowski NA, McClish DK. Design of Diagnostic Accuracy Studies. Statistical Methods in Diagnostic Medicine: John Wiley & Sons, Inc.; p Adibe OO, Amin SR, Hansen EN, Chong AJ, Perger L, Keijzer R, et al. An evidence-based clinical protocol for diagnosis of acute appendicitis decreased the use of computed tomography in children. Journal of Pediatric Surgery. 2011;46(1): Bergeron E, Richer B, Gharib R, Giard A. Appendicitis is a place for clinical judgement. The American journal of surgery. 1999;177(6): Ball CG, Dixon E, MacLean AR, Kaplan GG, Nicholson L, Sutherland FR. The impact of an acute care surgery clinical care pathway for suspected appendicitis on the use of CT in the emergency department. Can J Surg. 2014;57(3): Bakeman A, Somers KK, Browne L, Ertl A, Schultz J, Cassidy L, et al. Appendicitis scoring system decreases the incidence of "missed appendicitis". Journal of the American College of Surgeons. 2014;1):S Campbell H, Hotchkiss R, Bradshaw N, Porteous M. Integrated care pathways. BMJ: British Medical Journal. 1998;316(7125):

90 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain 76

91 Appendix 1: The APPEND scoring system Appendix 1: The APPEND scoring system APPEND SCORE Relative Anorexia = 1 point 1 Migratory Pain = 1 point 1 Localised Peritonism = 1 point 1 Clinical Management Score of 0-1 = Low suspicion of appendicitis. Discharge with GP follow up Score of 2-4 = Observe overnight. Consider USS for patients under 50, CT for patient over 50 Score of 5-6 = high risk of appendicitis. Consider diagnostic laparoscopy Elevated CRP >15 = 1 point 1 Elevate Neutrophils > 7.5 = 1 point 1 Dude (male gender) = 1 point 1 Total score out of 6 points 6 77

92 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain 78

93 Appendix 2: Staff satisfaction survey Appendix 2: Staff satisfaction survey 79

94 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain 80

95 Appendix 3: Sticker given to admitting staff Appendix 3: Sticker given to admitting staff Crit eria H I S T O R Y E X A M RIF/RLQ pain, Appendix intact, no generalised YES NO peritonism, current episode of Pain < 7d Anorexia YES NO Initial location Duration of pain (days) DAYS Worse with movement/cough YES NO Nausea YES NO Vomiting YES NO Diarrhoea YES NO Distension YES NO Migratory Pain YES NO Dysuria YES NO Days since last LMP Recent change in partner/ unprotected sex YES NO Change in nature of vaginal discharge YES NO Dyschezia YES NO Dyspareunia YES NO Dysmenorrhoea YES NO Previous abdominal surgery YES NO Temp 0 C Localised Peritonsim in RIF (percussion YES NO tenderness, guarding) Distension YES NO Generalised peritonism YES NO Rovsing s positive YES NO Ix Neutrophillia (>7.5) YES NO CRP (>15) YES NO MSU CHECKED YES NO APPEND SCORE (MALE +1) /out of 6 81

96 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain 82

97 Appendix 4: Spreadsheet used to record findings Appendix 4: Spreadsheet used to record findings DEMOGRAPHICS NHI Age Gender Ethnicity 1 BPZ ATF RZB other european NZ european other european 4 CRG NZ maori 5 RXK Fijian 6 GLN HNV NZ european NZ european 8 JGE INIDIAN 9 AEA RYR MZT cook Island Maori Other Asian Nz european 12 GLP Samoan 13 LWN NZ european 14 CEH NZ european 15 VFE Chinese Admission date 15/12/2016 2:28 15/12/ :45 15/12/ :31 15/12/ :20 16/12/ :40 18/12/ :35 18/12/ :29 19/12/ :07 17/12/ :55 20/12/ :36 21/12/2015 8:17 21/12/ :11 21/12/ :40 22/12/ :05 22/12/ :17 Discharge date 17/12/ :58 17/12/ :50 15/12/ :04 17/12/ :40 19/12/ :21 19/12/ :05 20/12/ :33 20/12/ :30 20/12/ :50 22/12/ :20 23/12/2015 9:49 21/12/ :11 28/12/2015 8:16 23/12/15 18:00 (no d/c or episode list) 23/12/ :58 Length of stay Admit to XR Admit to USS Admit to CT x 25 x x x x x 4 49 x x (CXR) 8 (AXR) x

98 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain HISTORY Imaging to OT Admit to OT OT to home Consultant Team Pain duration Migratory pain Anorexia Nausea Vomiting Dysuria Fevers Diarrhoea 1 Biggar RUB rahman RUB rahman RUB rahman RUB Wagener CBW x x x Moss PM x x x moss PM x 4 20 Moss PM x x x gynae gynae peng HPW Israel CI ED (gen surg review) ED Israel CI Biggar RUB Biggar RUB

99 Appendix 4: Spreadsheet used to record findings PV bleed PV d/c HISTORY EXAMINATION FINDINGS Dysmenorrhoea Dyspareunia Recent partner change Current time in cycle RLQ tenderness Localised peritonism Rovsing Hernia week post LMP Cx of hernia Testes exam x x x x x x mirena post mena

100 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain Testes exam findings EXAMINATION FINDINGS INITIAL OBSERVATIONS PV d/c Tender adnexa Cx excitation Adnexal mass Temp Febrile HR Systolic BP Diastolic BP RR O2 sats NAD

101 Appendix 4: Spreadsheet used to record findings WBC Leucocytosis Neut Left shift INVESTIGATIONS Lymph CRP Platelets Bili B- HcG MSU findings HV swab findings Endo swab findings x 0 x x x CXR findings x x x x 0 x x x x 0 x x x x x 0 x x x <2 1 x x x x 0 x x x x NAD NAD NAD x 0 x x NAD x 0 x x x x 0 x x x < x x NAD

102 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain INVESTIGATIONS AXR findings XR date USS USS findings TV USS findings USS date CT CT findings Colon Colon findings 1 x x 1 gallstones x x 1 no renal stones 2 x x x x x x x x x x 3 x x x x x x x x x x 4 x x x x x x /12/2015, 17/12/ x x 1 early appendicitis with possible reactive illeal changes x x x NAD, appendix not seen x x x x x x 7 x x 1 UTI x x x x x x 8 x x x x x x x x x x 9 NAD 18/12/ R ovary 36.3 ml 1 bulky R ovary x x x x 10 x 20/12/2015 x x x x 1 Acute appendicitis x x 11 x x x x x x 1 acute retrocecal appendicitis x x 12 x x x x x x x x x x 13 x x x x x x 1 Cecal volvulus, Bosniak L kidney cyst x x pulmonary nodule, no intra-abdo mm right proximal ureteric partially obstructing calculus with mild hydroureter and hydronephrosis. 88

103 Appendix 4: Spreadsheet used to record findings Alvarado score Initial APPEND 1 3 Evol Append Preop abx POST OP ANTIBIOTICS DURATION Surgical procedure performed Surg date ASA SURGICAL PROCEDURES Procedure code Procedure 1 Procedure hours lap 15/12/ Dx Lap Lap Appendicectomy open app 17/12/ hrs Lap appendicectomy 19/12/ laparoscopic appendicectomy 21/12/ laparoscopic appendicectomy la app 22/02/ R hemi 22/12/2015 2E

104 Prospective validation of the APPEND clinical prediction rule within a pathway dedicated to right iliac fossa pain 1 Procedure 3 Procedure 4 Gynae intraop Intraop findings Intra-op Dx 2 1 appendicitis tip appendicitis Intraop Dx 2 Histology acute suppurative appendicitis Serositis, no definitive evidence of inflammation arising within appendicitis 5 2 appendicitis gangrenous appendicitis acute appendicitis acute appendicitis inflammation no perforation acute appendicitis with perforation + appendiceal diverticulum 11 1 perforated perforated appendicitis x x Grade of appendicitis 90

105 Appendix 4: Spreadsheet used to record findings Cx Grade of cx COMPLICATIONS Major vs minor Cx Readmission Final diagnosis 1 multiple comorbidities, no appendicitis 2 appendicitis 3 Mesenteric adenitis 4 serositis 5 appendicitis 6 mesenteric adenitis 7 UTI 8 Appendicitis 9 PID perforated appendicitis perforated appendicitis alcoholic gastritis + vomiting 13 caecal volvulus 14?resolved renal colic- no d/c summary 15 Renal colic Secondary diagnosis Comments represented with RUQ pain - gastroscope duodenal polyp admitted earlier to ED and discharged - no append score calculated MSU not checked prior to USS! No discharge summary available 91