Factors contributing to the time taken to consult with symptoms of lung cancer: a cross-sectional study

Transcription

1 1 Centre of Aademi Primary Care, University of Aberdeen, Aberdeen, UK; 2 Department of General Pratie and Primary Care, University of Glasgow, Glasgow, UK; 3 Department of Publi Health, University of Aberdeen, Aberdeen, UK; 4 Department of Nursing and Midwifery, University of Stirling, Stirling, UK; 5 Department of Primary Health Care, University of Oxford, Oxford, UK; 6 Aberdeen Royal Infirmary, NHS Grampian, Aberdeen, UK Correspondene to: Ms S M Smith, Centre of Aademi Primary Care, University of Aberdeen, Foresterhill Health Centre, Westburn Road, Aberdeen AB25 2AY, UK; s.m.smith@abdn.a.uk Reeived 30 January 2008 Aepted 6 November 2008 Published Online First 2 Deember 2008 Fators ontributing to the time taken to onsult with symptoms of lung aner: a ross-setional study S M Smith, 1 N C Campbell, 1 U MaLeod, 2 A J Lee, 1 A Raja, 3 S Wyke, 4 S B Ziebland, 5 E M Duff, 2 L D Rithie, 1 M C Niolson 6 ABSTRACT Objetives: To determine what fators are assoiated with the time people take to onsult with symptoms of lung aner, with a fous on those from rural and soially deprived areas. Methods: A ross-setional quantitative interview survey was performed of 360 patients with newly diagnosed primary lung aner in three Sottish hospitals (two in Glasgow, one in NE Sotland). Supplementary data were obtained from medial ase notes. The main outome measures were the number of days from (1) the date partiipant defined first symptom until date of presentation to a medial pratitioner; and (2) the date of earliest symptom from a symptom heklist (derived from linial guidelines) until date of presentation to a medial pratitioner. Results: 179 partiipants (50%) had symptoms for more than 14 weeks before presenting to a medial pratitioner (median 99 days; interquartile range ). 270 partiipants (75%) had unreognised symptoms of lung aner. There were no signifiant differenes in time taken to onsult with symptoms of lung aner between rural and/or deprived partiipants ompared with urban and/or affluent partiipants. Fators independently assoiated with inreased time before onsulting about symptoms were living alone, a history of hroni obstrutive pulmonary disease (COPD) and longer pak years of smoking. Haemoptysis, new onset of shortness of breath, ough and loss of appetite were signifiantly assoiated with earlier onsulting, as were a history of hest infetion and renal failure. Conlusion: For many people with lung aner, regardless of loation and soioeonomi status, the time between symptom onset and onsultation was long enough to plausibly affet prognosis. Long-term smokers, those with COPD and/or those living alone are at partiular risk of taking longer to onsult with symptoms of lung aner and pratitioners should be alert to this. Only 7% of patients with lung aner in Sotland are alive 5 years after diagnosis; 1 a similar piture exists in England and Wales. 2 Five-year lung aner survival rates in the UK are well below the average rates of Europe and Ameria. 2 So too are surgial resetion rates; only 11% of patients in the UK reeive urative surgery ompared with 17% in Europe and 21% in North Ameria. 1 Differenes in olletion and presentation of data may aount for some of this variation, but the onsisteny of international omparisons suggests that other fators are also at play. 3 Stage at diagnosis is one suh fator. Most patients with lung aner in the UK are diagnosed at a late stage when urative surgery is no longer an option. 4 5 Late presentation Lung aner of symptoms to a medial pratitioner is reognised as being part of the problem Late presentation of aner symptoms is a ommon and well-doumented event Deisions on when to seek medial help are often bound up in a omplex mix of individual psyhosoial fators The effets of previous publi and professional neglet of lung aner may also still endure. A reent qualitative study reported late presentation of lung aner symptoms as a universal feature. 8 Partiipants, regardless of their disease stage, failed to reognise the seriousness of their symptoms and reported notieable hanges in health long before seeking medial help. This and other lung aner researh 15 hallenges the general belief that lung aner is asymptomati until advaned. Geographial variation in lung aner survival also exists within the UK. Researh into aner inequalities in Sotland has shown that people from rural and deprived areas have poorer lung aner survival than people from urban and affluent areas There is no evidene that healthare providers in Sotland differentially delay referral and/or treatment for lung and other aners in these groups Differenes in attitude to onsulting a liniian with symptoms of aner between people in rural and urban areas have been reported, 23 but whether those in rural and deprived ommunities wait longer before seeking medial help is unknown. A delay of even a few months in seeking medial help is believed to have a signifiant influene on stage at diagnosis and prognosis. 24 Reduing the time taken to onsult with symptoms of lung aner in those more likely to postpone seeking medial help may help improve survival. 25 However, as the reasons for late presentation are not fully understood, a greater appreiation of the proess that ours before first ontat with a medial pratitioner is required. This study aimed to explore this proess and to determine what patient fators are assoiated with the time taken to onsult with symptoms of lung aner, and whether those from rural and/or deprived areas wait longer than those from urban and/or affluent areas before seeking medial help. METHODS Conseutive patients with primary lung aner were identified by speialist nurses at three Sottish hospitals (one in north-east Sotland and two in Glasgow) early after diagnosis between April 2004 and January Both loations treat patients aross the spetrum of rurality and deprivation. Thorax 2009;64: doi: /thx

2 Table 1 Partiipant demographi and soioeonomi harateristis (n = 360) Charateristi No (%) Age group (years) (49 18 (5) (21) (32) (34) >80 32 (9) Sex Male 210 (58) Female 150 (42) Ethniity White 359 (.99) Pakistani 1 (,1) Eduation No qualifiations 153 (43) Higher/A level 110 (31) HNC/HND 10 (3) Professional/tehnial qualifiations 54 (15) University degree 15 (4) GSVQ/SVQ 18 (5) Smoking Ex-smoker 235 (65) Smoker at time of interview 110 (31) Never smoker 15 (4) Area North-east Sotland 190 (53) Glasgow 170 (47) Home ownership Own home 218 (60) Rent home 132 (37) Other 10 (3) Travelling distane to nearest aner entre (miles),5 187 (52) 5,10 51 (14) 10,15 14 (4) 15,20 17 (5) 20,25 5 (1) 25,30 9 (3) >30 77 (21) Travelling distane to GP (miles) (1 120 (33) 1,3 163 (45) 3,5 56 (15) 5,7 11 (4) 7,11 10 (3) Indiators of deprivation were assigned to ases aording to their postode at the time of diagnosis. Carstairs deprivation sores were alulated from 2001 ensus data 26 at the output area level (the lowest level geography used to produe statistial output from the ensus) and grouped into population quintiles. Rural and urban status was alloated aording to quintiles of travelling distane from a partiipant s postode to the nearest aner entre; this measure was previously found to be assoiated with diagnosis of lung aner at a later stage and poorer early survival. Beause of diffiulties in answering survey questions, patients who had impaired onsiousness, dementia or psyhosis were exluded from the study. Patients onsidered by liniians to be too unwell to be approahed were also exluded. Partiipants agreeable to the study reeived an information sheet and allowed their details to be passed to the study researher. Contat was made and the study was explained in full before obtaining signed onsent. Data were then olleted using a quantitative interview survey (see below). In addition, dates of onsultations, referral and first hospital ontat, and data on o-morbidities (previous aner, hest and ardiovasular disease, gastrointestinal and hepati disease, musuloskeletal problems and psyhiatri problems) were abstrated from GP and hospital ase notes. Quantitative interview survey Partiipants were asked about their initial symptoms and the dates they were first notied. Two definitions of first symptom were used partiipant-defined and health professionaldefined using a heklist of symptoms ompiled from Caner Researh UK lung aner symptoms 27 and SIGN guidelines. 1 Further data inluded knowledge of lung aner symptoms; previous onsulting behaviour; exposure to risk fators; family history of aner; and pereption of self-risk of aner. Individual soioeonomi data (inluding eduation, employment and home ownership) were olleted to test deprivation validity at the individual patient level. Additional variables shown to be assoiated with illness behaviour suh as living alone, having someone with whom to disuss symptoms and relationship with their GP were also inluded. Outome measures The main outomes were (1) the number of days from date of first symptom defined by the partiipant until date of presentation of symptoms to a medial pratitioner and (2) the number of days from date of earliest symptom from the symptom heklist until date of presentation of symptoms to a medial pratitioner. Study power A total sample size of 400 was projeted to give 80% power to detet an absolute differene of 14% (50% versus 36%) in two equal group omparisons of patients waiting.12 weeks before presenting to a medial pratitioner at the 5% signifiane level. If numbers in one group fell to 150, the study would have a similar power to detet a differene of 16%. Statistial analysis Statistial analyses were onduted using SPSS for Windows Version 15. Univariate analyses (x 2 tests, Pearson orrelation and t tests) were performed to examine the assoiation of the independent variables with eah outome variable. Time to onsultation had a skewed distribution and a log transformation was required before analysis. The results were onverted bak to geometri means (transformed 95% onfidene intervals, CI) for presentation of the univariate results. To determine whih partiipant fators had an independent assoiation with eah outome, multiple linear regression analyses were performed. The fators age, sex, distane to aner entre, Carstairs deprivation sores and area (north-east Sotland or Glasgow) were initially fored into the model. Other fators were entered into the initial model if they had shown a signifiant univariate assoiation at p(0.20. A bakward stepwise removal proess seleted preditor variables if p In order to make the interpretation of the regression models easier, the antilog of eah regression oeffiient and its 95% CI are presented. For ategorial fators the resultant values an be interpreted as the number of times longer that a partiular group took to onsult their GP ompared with the referene group. For ontinuous 524 Thorax 2009;64: doi: /thx

3 Table 2 Earliest heklist symptoms experiened, partiipant-defined first symptoms and heklist symptoms at time of presentation to a medial pratitioner (n = 360) Symptoms fators the oeffiients refer to the proportionate hange in time taken to onsult for eah unit inrease in the ontinuous fator. RESULTS Over the 92-week study period, 835 people with primary lung aner were identified; 154 people were ineligible, most of whom were deemed too ill by liniians to be approahed about the study. A further 23 whose eligibility status was not onfirmed had also to be exluded. A total of 658 were eligible for the study and 620 were approahed; 113 of these refused. Of the remaining 507, 96 withdrew before giving full onsent, 24 died before an interview ould take plae and ontat was lost with 26. A total of 361 partiipants were reruited and ompleted quantitative interview surveys (all were fae-to-fae apart from two whih were onduted by telephone). One partiipant reported first presenting with symptoms of lung aner 10 years previously and was exluded from the rest of the analyses as an outlier. Three hundred and forty-six (96%) of both hospital and GP ase notes were reviewed. Comparing dates of first onsultation reported by partiipants with GP ase note reorded dates, 32% (n = 102) were within 1 week, 36% (n = 114) between 8 and 30 days and 32% (n = 100) more than 1 month. No evidene of reporting bias between affluent, deprived, rural and urban groups was found. Partiipant harateristis The sample (n = 360) onsisted of 58% men and 42% women. The harateristis of the partiipants are shown in table 1. Their median age was 68 years (interquartile range (IQR) 59 74), ranging from 37 to 87 years. Sixty-two (17%) lived in the least deprived areas (Carstairs quintile 1) and 128 (36%) lived in the most deprived areas (Carstairs quintile 5). The median travelling distane to a aner entre was 4.5 miles (IQR ), range miles, and 77 (21%) lived 30 miles or more from their nearest aner entre. Three hundred and thirty-five (78%) lived within 3 miles of their GP. Only 4% of the partiipants had never smoked and 201 (56%) of smokers and ex-smokers reported not pereiving themselves to be at risk of lung aner before their diagnosis despite being aware of the risks of smoking. Two hundred and seventy (75%) reported having Earliest heklist symptom(s) experiened No (%) Partiipant-defined first symptom(s) No (%) Cough 70 (19) 117 (32) 218 (61) Shortness of breath 76 (21) 81 (22) 211 (59) Coughing up phlegm 49 (14) 27 (7) 177 (49) Tiredness/fatigue/lethargy 57 (16) 51 (14) 118 (33) More short of breath than usual 41 (11) 11 (3) 126 (35) Loss of weight 71 (20) 27 (7) 124 (34) Loss of appetite 57 (16) 23 (6) 111 (31) Pain in hest/shoulders/bak related to 49 (14) 53 (15) 108 (30) breathing or oughing Hoarse voie 38 (10) 8 (2) 81 (22) Change in a ough 18 (5) 3 (1) 68 (19) Haemoptysis 17 (5) 44 (12) 61 (17) Diffiulty swallowing 12 (3) 5 (1) 23 (6) Swelling of the fae or nek 11 (3) 11 (3) 21 (6) No symptoms (eg, inidental finding) 27 (7) 4 (1) Median (IQR) number of symptoms experiened 1 (1 2) 2 (1 2) 4 (2 5) Cheklist symptom(s) at time of presentation No (%) no knowledge of lung aner symptoms and 171 (51%) reported not having believed that their first symptom(s) was serious. Time to presentation of symptoms The median time from partiipant-defined first symptoms to onsultation with a medial pratitioner was 21 days (IQR 7 56). The median time from earliest reported heklist symptoms until onsultation was 99 days (IQR ). Reognition of symptoms Most partiipants, even those who desribed themselves as having no first symptoms, reported additional symptoms in response to the symptom heklist, suggesting that these symptoms were possibly ignored or not reognised as relevant, salient or indiative of serious illness (table 2). Effets of deprivation and rurality On univariate analysis, neither deprivation nor rurality was signifiantly assoiated with time to onsultation. This was onfirmed after adjustment for other signifiant variables in both models (tables 3 and 4). Of the related variables, inreasing travel distane to general praties (p = 0.079) and soial ontat with GPs away from the surgery (p = 0.049), both of whih are more ommon in rural areas, were assoiated with inreasing time between first partiipant-defined symptom and onsultation on univariate analysis, but not when adjusted for other fators (table 3). People who were urrently or had previously been in paid employment onsulted sooner after their earliest heklist symptom than those who had never been in paid employment (p = 0.037), but the differene was no longer statistially signifiant after adjusting for other variables (p = 0.092, table 4). A wide variety of fators were assoiated univariately with time to onsultation. However, different fators were independently related to the time between first heklist symptom and presentation and the time between first partiipant-defined symptom and presentation (tables 3 and 4). Thorax 2009;64: doi: /thx

4 Table 3 Univariate/multivariate analyses of fators assoiated with partiipant-defined first symptoms until onsultation with medial pratitioner Univariate analysis Multivariate analysis* Categorial variables No Geometri mean (95% transformed CI) p Value Regression oeffiient antilog{ (95% CI) p Value First symptoms experiened Cough No (14.6 to 21.7) Yes (17.9 to 32.2) Change in ough No (15.0 to 21.1) Yes (21.8 to 83.4) Loss of appetite No (16.6 to 24.0) Yes (9.8 to 21.4) Co-morbidities History of asthma No (16.4 to 23.2) Yes (7.1 to 23.2) History of hest infetion No (16.3 to 23.3) Yes (5.6 to 18.7) History IHD, stroke, heart failure No (18.6 to 26.9) Yes (7.5 to 16.1) 0.57 (0.38 to 0.83) History of hepati disease No (16.4 to 23.2) Yes (5.0 to 24.1) History of upper GI disease No (13.9 to 20.8) Yes (16.5 to 31.8) 1.62 (1.12 to 2.40) History of renal failure No (16.0 to 22.7) Yes (4.4 to 26.3) History of alohol/drug misuse No (16.3 to 23.4) Yes (7.7 to 22.5) Demographi/soioeonomi Sex Male (13.6 to 21.2) (0.49 to 0.98) Female (16.5 to 27.7) 1.00 Area North-east Sotland (17.6 to 27.1) Glasgow (12.2 to 20.5) 0.84 (0.54 to 1.32) Carstairs (deprived) sore in quintiles 1 (least (13.9 to 33.0) (0.85 to 1.17){ (15.9 to 31.6) (13.3 to 31.4) (8.8 to 22.6) 5 (most) (12.4 to 22.1) Distane between aner entre and partiipant 1 (nearest) (12.9 to 26.9) postode address (quintiles) (10.3 to 22.3) 1.04 (0.89 to 1.20){ (12.2 to 27.1) (12.2 to 27.1) 5 (furthest) (18.1 to 33.2) Continued 526 Thorax 2009;64: doi: /thx

5 Table 3 Continued Univariate analysis Multivariate analysis* Categorial variables No Geometri mean (95% transformed CI) p Value Regression oeffiient antilog{ (95% CI) p Value GP onsulting/relationship: Before your diagnosis, how many times a year did you see your GP What would you do if you were not happy with the quality of servie your GP pratie provides? Do you know your usual GP away from the surgery? None (22.8 to 48.1) One (14.4 to 25.2) 0.60 (0.37 to 0.95) times (11.5 to 21.8) 0.47 (0.27 to 0.81) times (9.0 to 18.5) 0.41 (0.23 to 0.72) Do nothing (14.1 to 30.9) Complain to GP (15.3 to 27.2) Complain to higher authority (23.0 to 95.3) Change GP pratie (12.0 to 22.9) Don t know (8.8 to 19.8) Not at all/not very well (14.4 to 20.9) Fairly/very/extremely well (19.2 to 45.5) Smoking history: Are you a smoker? Never smoker (15.7 to 73.3) Current smoker (11.3 to 20.3) Ex-smoker (16.2 to 24.6) Symptom knowledge Knowledge of lung aner symptoms before diagnosis Nothing (13.9 to 20.7) A little (13.8 to 30.2) 1.07 (0.69 to 1.66) Quite a lot/everything (33.4 to 68.6) 2.43 (1.26 to 4.79) Continuous variables Correlation oeffiient p Value B p Value Age Per year (0.98 to 1.02) Travelling distane to GP Per mile *Model inluded those ategorial and ontinuous variables that were signifiant on univariate analyses at the p(0.20 level. {The regression oeffiient antilog an be interpreted as (1) for ategorial fators: the number of times longer for partiular groups to onsult with a medial pratitioner; and (2) for ontinuous fators: the proportionate hange in time to onsult per unit inrease. {Linear trend aross ategories. GP, general pratitioner; GI, gastrointestinal; IHD, ishaemi heart disease. Thorax 2009;64: doi: /thx

6 Table 4 Univariate/multivariate analyses of fators assoiated with time from earliest heklist symptom until onsultation with medial pratitioner Univariate analysis Multivariate analysis* Categorial variables No Geometri mean (95% transformed CI) p Value Regression oeffiient antilog{ (95% CI) p Value Earliest symptoms experiened Cough No (96.1 to 150.4) Yes (30.5 to 65.5) 0.54 (0.33 to 0.85) Change in ough No (81.3 to 123.0) Yes (25.1 to 100.1) Shortness of breath No (93.7 to 147.4) Yes (33.6 to 71.5) 0.44 (0.28 to 0.69) Coughing up phlegm No (90.5 to 139.2) Yes (25.3 to 65.6) Coughing up phlegm with signs of blood No (83.7 to 125.2) Yes (11.3 to 66.8) 0.29 (0.12 to 0.66) Loss of appetite No (83.9 to 130.6) Yes (41.5 to 100.5) 0.48 (0.30 to 0.79) Hoarse voie No (73.8 to 112.7) Yes (82.4 to 277.9) Swelling in fae or nek No (80.0 to 119.3) Yes (10.5 to 164.8) Co-morbidities History of COPD No (65.2 to 101.4) Yes (105.7 to 260.6) 1.86 (1.17 to 2.88) History of hest infetion No (83.3 to 125.6) Yes (19.1 to 144.1) 0.35 (0.17 to 0.74) History of osteoporosis No (75.5 to 115.4) Yes (90.8 to 317.5) History of lower GI disease No (83.4 to 125.6) Yes (24.2 to 143.2) History of upper GI disease No (69.9 to 113.6) Yes (84.3 to 179.5) History of renal failure No (83.5 to 125.3) Yes (8.3 to 115.3) 0.38 (0.16 to 0.91) Demographi/soioeonomi Sex Male (77.1 to 129.8) (0.13 to 1.78) Female (68.5 to 127.1) 1.00 Area North-east Sotland (78.5 to 133.3) Glasgow (65.8 to 120.6) 0.86 (0.52 to 1.41) Distane between aner entre and partiipant 1 (nearest) (55.4 to 124.8) (0.89 to 6.31){ postode address (quintiles) (70.9 to 169.7) (58.4 to 156.3) (64.7 to 169.5) 5 (farthest) (59.8 to 139.1) Continued 528 Thorax 2009;64: doi: /thx

7 Table 4 Continued Univariate analysis Multivariate analysis* Categorial variables No Geometri mean (95% transformed CI) p Value Regression oeffiient antilog{ (95% CI) p Value Carstairs (deprived) sore in quintiles 1 (least deprived) (52.0 to 145.8) (0.85 to 1.17){ (71.2 to 169.0) (67.8 to 177.2) (46.3 to 142.8) 5 (most deprived) (66.4 to 131.2) Employment status Others (61.0 to 108.3) Retired (85.3 to 147.5) Have you ever had paid work? No (86.6 to 138.8) Yes (47.0 to 97.3) 0.65 (0.39 to 1.07) Risk fators Did you pereive yourself to be at any risk from Low risk/not at all (68.7 to 110.4) lung aner before your diagnosis? Average risk (42.5 to 112.7) Fairly/very high risk (111.4 to 311.4) Have you ever worked in a smoky environment? No (58.3 to 108.3) Yes (84.3 to 142.6) Soial fators Do you live? With others (64.9 to 101.8) On my own (103.1 to 232.6) 1.99 (1.29 to 3.09) Continuous variables Correlation oeffiient p Value B (95% CI for B) p Value Pak years of smoking (1.02 to 1.20) Age Per year (0.98 to 1.02) *Model inluded those ategorial and ontinuous variables that were signifiant on univariate analyses at the p(0.20 level. {The regression oeffiient antilog an be interpreted as (1) for ategorial fators: the number of times longer for partiular groups to onsult with a medial pratitioner and (2) for ontinuous fators: the proportionate hange in time to onsult per unit inrease. {Linear trend aross ategories. COPD, hroni obstrutive pulmonary disease; GI, gastrointestinal. Thorax 2009;64: doi: /thx

8 The shorter time between partiipant-defined first symptom and presentation to a medial pratitioner showed little relationship with symptoms. On the other hand, it was predited by previous illness, being shorter in those with ardiovasular disease but longer in those with previous upper gastrointestinal disease. People who regularly onsulted their GP before their diagnosis onsulted sooner with lung aner symptoms, and those who stated they would omplain to a higher authority if unhappy with their general pratie took longer (although this was not an independent preditor). For the longer time between earliest heklist symptom and presentation, more symptoms harateristis were signifiantly assoiated with time to onsultation (table 4). Of soial fators, living alone was the independent fator most strongly assoiated with time to onsultation. Related variables (eg, having someone to talk to about symptoms or being prompted to onsult by others) were not signifiant independent preditors. Some o-morbidities (previous hest infetion or renal failure) were independently assoiated with a shorter time to onsultation, while a history of hroni obstrutive pulmonary disease (COPD) and inreasing pak-years of smoking were independently assoiated with inreased time to onsultation. DISCUSSION We found no substantial differenes in the time taken to present with symptoms of lung aner between people living in rural, urban, affluent or deprived areas but, for all groups, these times were long. Most partiipants had unreognised symptoms of lung aner and half had experiened them for more than 14 weeks before presenting to a medial pratitioner, indiating that symptoms were either ignored or not understood to be potentially serious even after the diagnosis of aner had been made. This study is the largest of its kind to ollet suh extensive data on reported lung aner symptoms and symptom duration before onsultation that we have been able to identify. 28 This has allowed us to quantify the importane of symptoms, omorbidities and soial fators as determinants of this lag time. A limitation is that it relied on self-reporting by partiipants so may have been subjet to reall bias although, where we ould hek reall against ase note reords (for example, onsultation dates), we found it to be reasonably good. We need to aknowledge that reports of some behaviours are not just affeted by reall but by people s natural desire to avoid blaming themselves for waiting with serious symptoms or to avoid faing the stigma of a smoking-related disease. 29 Also, first symptoms of lung aner if non-speifi an easily be dismissed. We were also aware that reruiting patients with lung aner would be diffiult beause of their rapid linial deterioration and high early death rate, so monitored all losses at the various stages of reruitment. This proess onfirmed that a notable proportion died early or were otherwise medially unfit to take part, and our data may not be representative of these groups. Our findings add weight to qualitative researh that shows symptoms are experiened for several months before onsultation with a medial pratitioner, 8 even though it is generally thought to be asymptomati until it is advaned. This ontrasts with five previous quantitative studies whih reported median times to onsultation of 1 month or less. In the National Survey of NHS patients, 30 time before onsultation was only reported for the minority of patients who presented diretly to hospital either with inidental findings or as emergenies. This and the four smaller studies 28 did not use symptoms heklists and their findings are in line with the shorter lag time we found from patient-defined first symptoms: our researh shows that patients experiene unreognised symptoms for many preeding weeks. This study is also in agreement with other researh reporting on the most ommon first linial features of lung aner. 15 That smoking was assoiated with inreased time to onsultation might be viewed as ounterintuitive, given that the relationship between smoking and lung aner is so well known. On the other hand, we found that smokers said they did not pereive themselves to be at risk this unrealisti optimism has been reported in other studies of risk pereption, nor did we find any evidene that high pereived risk was assoiated with onsulting sooner (if anything, we found the reverse). It ould be that smokers were more tolerant of symptoms, regarding them as normal for smokers. The most important soial fator assoiated with time to onsult in our study was living alone, whih has been found previously in patients with breast aner and heart disease. This may be beause o-habitees notie symptoms and santion help-seeking behaviour. Previous researh has onentrated on the latter, 11 but our findings suggest the former may be more important. We have onfirmed the findings of reent qualitative researh whih suggested that o-morbidities may be important in ausing patients to hold off before onsulting, 8 and quantified their importane. In partiular, people with COPD took twie as long to onsult with symptoms. More enouragingly, we also found that some omorbidities were assoiated with redued time to onsultation; those who had previously experiened a hest infetion requiring hospital treatment onsulted in half the time of others. This suggests that people who have previously been alerted to ertain symptoms onsult with them more quikly, providing enouragement that effetive intervention may be possible. The impliations of our findings are twofold. First, lung aner is almost always symptomati, usually for several months before onsultation. The median time from the earliest symptom to onsultation is more than the estimated doubling time of lung aner and plausibly long enough to affet stage at diagnosis and prognosis. 35 It follows that, if people with symptoms onsulted earlier and were investigated appropriately, there is potential to improve survival from lung aner. Seond, most symptoms go unnotied for several weeks, so earlier onsultation will only our if people are alert to symptoms and know to onsult with them. Ahieving this will not, however, be easy. Key proesses inlude symptom reognition, motivation by people to have their symptoms investigated and stimulating triggers to onsultation. A lung aner awareness ampaign may be of merit. However, it is not yet known if this an be ahieved to an extent that ould affet survival, so further researh is needed. Further researh is also needed to understand the partiular issues relating to presentation in those who present in poor ondition or who rapidly deteriorate. Pratitioners, in partiular GPs, should be aware of the potential to diagnose lung aner earlier. This inludes reognising and ating on hest symptoms, partiularly in high-risk groups, and being aware that patients may not have understood the potential signifiane of their new hest symptoms, espeially if they have a history of COPD, live alone and have smoked for a long time. 530 Thorax 2009;64: doi: /thx

9 Aknowledgements: The authors thank all the patients who kindly gave their time to partiipate in the study and the following people for their help, advie and guidane throughout the study: the lung nurse speialists involved in reruitment (Iona Brisbane, Louise Brown, Penny Downer, Shona Haggart, Lynne MGuiness and John MPhelim); Dr David Dunlop, Consultant Medial Onologist; Dr Rihard Jones, Consultant Clinial Onologist; Dr Joe Legge, Consultant in Thorai Mediine; Dr Nazia Mohammed, Consultant Clinial Onologist; Dr Rob Milroy, Consultant in Respiratory Mediine; Dr Brian Neilly, Consultant Physiian; and Dr Marianne Niolson, Consultant Medial Onologist. Funding: This study was funded by Caner Researh UK. Their role was solely as a funding body. Competing interests: None. Ethis approval: This study was approved by the North of Sotland researh ethis ommittee and North Glasgow University Hospitals NHS Trust researh ethis ommittee. REFERENCES 1. Sottish Interollegiate Guidelines Network (SIGN). Management of lung aner. Edinburgh: SIGN, Caner Researh UK. Caner Help UK. CanerStats key fats on lung aner and smoking (updated 2007). #survival 3. Sottish Government. Caner in Sotland: sustaining hange (updated 21 June 2005) Sottish Exeutive Health Department. Caner senarios: an aid to planning aner servies in Sotland in the next deade. Edinburgh: The Sottish Exeutive, Caner Researh UK. Caner Help UK. Lung aner survival statistis (updated September 2007) Bowen EF, Rayner CF. Patient and GP led delays in the reognition of symptoms suggestive of lung aner. Lung Caner 2002;37: Kesson E, Buknall CE, MAlpine LG, et al. Lung aner management and outome in Glasgow, Br J Caner 1998;78: Corner J, Hopkinson J, Roffe L. Experiene of health hanges and reasons for delay in seeking are: a UK study of the months prior to the diagnosis of lung aner. So Si Med 2006;62: Burgess CC, Hunter MS, Ramirez AJ. A qualitative study of delay among women reporting symptoms of breast aner. Br J Gen Prat 2001;51: Tromp DM, Brouha XDR, De Leeuw JRJ, et al. Psyhologial fators and patient delay in patients with head and nek aner. Eur J Caner 2004;40: Smith LK, Pope C, Botha J. Patients help-seeking experienes and delay in aner presentation: a qualitative synthesis. Lanet 2005;366: Campbell SM, Roland MO. Why do people onsult the dotor? Fam Prat 1996;13: Andersen BL, Caioppo JT. Delay in seeking a aner diagnosis: delay stages and psyhophysiologial omparison proesses. Br J So Psyhol 1995;34: Loker D. Symptoms and illness. The ognitive organisation of disorder. London: Tavistok, Koyi H, Hillerdal G, Branden E. A prospetive study of a total material of lung aner from a ounty in Sweden : gender, symptoms, type, stage, and smoking habits. Lung Caner 2002;36: Campbell NC, Elliott AM, Sharp L, et al. Rural fators and survival from aner: analysis of Sottish aner registrations. Br J Caner 2000;82: MLaren G, Bain M. Deprivation and health in Sotland: insights from NHS data. Edinburgh: ISD Publiations, Campbell NC, Elliott AM, Sharp L, et al. Rural and urban differenes in stage at diagnosis of oloretal and lung aners. Br J Caner 2001;84: Pollok AM, Vikers N. Deprivation and emergeny admissions for aners of oloretum, lung, and breast in south east England: eologial study. BMJ 1998;317: Campbell NC, Elliott AM, Sharp L, et al. Impat of deprivation and rural residene on treatment of oloretal and lung aner. Br J Caner 2002;87: Maleod U, Ross S, Twelves C, et al. Primary and seondary are management of women with early breast aner from affluent and deprived areas: a retrospetive review of hospital and general pratie reords. BMJ 2000;320: Robertson R, Campbell NC, Smith S, et al. Fators influening time from presentation to treatment of oloretal and breast aner in urban and rural areas. Br J Caner 2004;90: Bain NSC, Campbell NC. Treating patients with oloretal aner in rural and urban areas: a qualitative study of the patients perspetive. Fam Prat 2000;17: Christensen ED, Harvald T, Jendresen M, et al. The impat of delayed diagnosis of lung aner on the stage at the time of operation. Eur J Cardiothora Surg 1997;12: Bozuk H, Martin C. Does treatment delay affet survival in non-small ell lung aner? A retrospetive analysis from a single UK entre. Lung Caner 2001;34: ISD Sotland. Information and statistis CRUK Caner Researh UK. Caner Help UK. Information servie about aner and aner are Madonald S, Maleod U, Mithell E, et al. Fators influening patient and primary are delay in the diagnosis of aner: a database of existing researh and its impliations for future pratie. Final Report to the Department of Health (Ref ). Glasgow: University of Glasgow, Chapple A, Ziebland S, MPherson A. Stigma, shame and blame experiened by patients with lung aner: qualitative study. BMJ 2004;328: Allgar VL, Neal RD. Delays in the diagnosis of six aners: analysis of data from the National Survey of NHS patients: Caner. Br J Caner 2005;92: Weinstein ND, Marus SE, Moser RP. Smokers unrealisti optimism about their risk. Tob Control 2005;14: Arnett J, Optimisti bias in adolesent and adult smokers and nonsmokers. Additive Behaviors 2000;25: Burgess CC, Ramirez AJ, Rihards MA, et al. Who and what influenes delayed presentation in breast aner? Br J Caner 1998;77: Horne R, James D, Petrie K, et al. Patients interpretation of symptoms as a ause of delay in reahing hospital during myoardial infartion. Heart 2000;83: Thunnissen FBJM, Shuurbiers OCJ, den Bakker MA. A ritial appraisal of prognosti and preditive fators for ommon lung aners. Histopathology 2006;48: Thorax: first published as /thx on 3 Deember Downloaded from BMJ Careers online re-launhes BMJ Careers online has re-launhed to give you an even better online experiene. You ll still find our online servies suh as jobs, ourses and areers advie, but now they re even easier to navigate and quiker to find. New features inlude: Job alerts you tell us how often you want to hear from us with either daily or weekly alerts Refined keyword searhing making it easier to find exatly what you want Contextual display when you searh for artiles or ourses we ll automatially display job adverts relevant to your searh Reruiter logos linked diretly to their organisation homepage find out more about the ompany before you apply RSS feeds now even easier to set up Visit areers.bmj.om to find out more. on 5 July 2018 by guest. Proteted by opyright. Thorax 2009;64: doi: /thx