Data from randomized controlled trials and observational

Transcription

1 GASTROENTEROLOGY 2011;140: CLINICAL An Alerting System Improves Adherence to Follow-up Recommendations From Colonoscopy Examinations DANIEL A. LEFFLER,* NAAMA NEEMAN, JAMES M. RABB,* JACOB Y. SHIN, BRUCE E. LANDON, KUMAR PALLAV,* Z. MYRON FALCHUK,* and MARK D. ARONSON *Department of Gastroenterology, Department of Medicine, Beth Israel Deaconess Medical Center, Boston, Massachusetts; Department of Health Care Policy, Harvard Medical School, Boston, Massachusetts This article has an accompanying continuing medical education activity on page e13. Learning Objective: Upon completion of this CME module successful learners will be able to explain issues in follow-up colonoscopy adherence and identify potential solutions to improve follow up and reduce risk. See related article, Hjelkrem M et al, on page 326 in CGH; see editorial on page BACKGROUND & AIMS: Systems are available to ensure that results of tests are communicated to patients. However, lack of adherence to recommended follow-up evaluation increases risk for adverse health outcomes and medical or legal issues. We tested the effectiveness of a novel follow-up system for patients due for surveillance colonoscopy examinations. METHODS: Electronic medical records from colonoscopies performed 5 years prior were reviewed to identify individuals due for a repeat surveillance colonoscopy examination. Patients were assigned to groups that received the standard of care or a newly developed follow-up system that included a letter to the primary care provider, 2 letters to the patient, and a telephone call to patients who had not yet scheduled an examination by the procedure due date. The primary end point was the percentage of patients who scheduled or completed the colonoscopy examination within 6 months of the due date. Secondary end points included detection rate for adenomas, sex- and ethnicity-specific follow-up rates, and patient satisfaction. RESULTS: Of 2609 patient records reviewed, 830 (31.8%) were found to be due for a surveillance colonoscopy examination in the study period. At the conclusion of the study, 241 (44.7%) patients in the intervention had procedures scheduled or completed, compared with 66 (22.6%) in the control group (P.0001). The follow-up system appeared particularly effective among non-white patients; patients reported general satisfaction with the reminder program. CONCLUSIONS: A simple protocol of letters and a telephone call to patients who are due for colonoscopy examinations significantly improved adherence to endoscopic follow-up recommendations. This work provides justification for the creation of reminder systems to improve patient adherence to medical recommendations. Keywords: Quality Improvement; Screening; Colon Cancer Surveillance; Efficacy. Data from randomized controlled trials and observational studies show that colonoscopy is associated with a reduction in both colorectal cancer incidence and mortality. 1,2 Colorectal cancer arises from a series of sequential genetic mutations that result first in adenomatous polyps, which then progress on to the development of cancer. As the only screening method that directly visualizes the entire colon, colonoscopy has become a preferred method of colorectal cancer screening in the United States. 3 Although evidence-based guidelines for colon cancer screening and surveillance exist, 4 there are significant issues with patient adherence to recommendations regarding colonoscopy. Many studies document barriers to initial colonoscopy, 5 11 but even after the initial colonoscopy is performed, appropriate follow-up testing often is neglected with potentially serious consequences. Limited data suggest that between 15% and 30% of patients, including those with a history of polyps, do not adhere to surveillance colonoscopy recommendations, although these studies were in selected groups and likely were not representative of a typical colonoscopy surveillance pop- Abbreviations used in this paper: BIDMC, Beth Israel Deaconess Medical Center; EMR, electronic medical record; PCP, primary care physician by the AGA Institute /$36.00 doi: /j.gastro

2 April 2011 ALERTING SYSTEM 1167 ulation. 12,13 Many practices and institutions currently have systems in place to ensure and document that results of tests, such as pathology findings after a polypectomy, are communicated to patients. Few institutions or practices, however, have implemented systems to monitor and improve compliance with suggested follow-up tests. Reasons for the current lack of systems to improve adherence with endoscopic follow-up evaluation include the length of time between procedures, controversy over optimal follow-up periods, and uncertainty regarding specialist vs primary care physician responsibility. Regardless of cause, the lack of system-wide strategies for ensuring recommended follow-up evaluation after procedures carries with it many important clinical ramifications. Most importantly, patient care is compromised when delayed or missed procedures lead to the development of more advanced disease, as occurs when advanced lesions are found. In addition, there is an increased risk of litigation when there is no reminder system in place to inform patients of the need for repeat studies, and no documentation that patients have received adequate notification of the need for follow-up testing. Data from a large malpractice carrier show that missed diagnosis of colorectal cancer is the second most common type of cancer cited in malpractice claims, and that many of these claims were related to the lack of adequate follow-up evaluation and documentation systems. 14 Although all follow-up medical tests are subject to variable adherence rates, colonoscopy presents particular difficulty for both patients and providers because of variability in the recommended follow-up interval and long length of time between examinations. To address the earlier-mentioned barriers for effective endoscopic follow-up evaluation, we developed, implemented, and tested a comprehensive follow-up management system for incorporation into the electronic medical record (EMR) of a major gastroenterology referral center that performs approximately 10,000 screening colonoscopies per year. This article describes the specifications of the follow-up system created and the results of a randomized controlled trial of the effectiveness of this system. Materials and Methods Development of Principles for an Automated Follow-up Reminder System Interviews and meetings with primary care and gastroenterology physicians suggested overlapping but not identical requirements for a colonoscopy follow-up management system that included the following 7 principles: (1) both the specialist and primary care physician (PCP) should be able to place and alter the colonoscopy referral; (2) PCPs should be notified when the follow-up procedure is due, and before the patient is contacted so that, when appropriate, the order can be modified or canceled; (3) for efficiency, patients first should be contacted by letters, with telephone contact reserved for necessary cases; (4) PCPs should be notified if a patient fails to keep recommended follow-up testing despite the reminders; (5) specialists should be able to choose on a case-by-case basis to be notified if the patient does not schedule/keep a follow-up appointment (eg, for follow-up evaluation of high-risk patients); (6) the reminder system should be patient-dependent rather than provider-dependent so that the system will remain functional regardless of whether the initial or ordering physician is still involved in caring for the patient; and (7) the reminder system should be integrated into the EMR in a manner that minimizes change to physician workflow. Design of the Automated Follow-up Reminder System In accordance with the 7 core principles listed earlier, we developed a system algorithm wherein ordering prompts are linked to the online pathology reports, to encourage specialists to order or modify follow-up recommendations as necessary based on the pathology results. Links to current guidelines also are embedded in the ordering system. The system is initiated when a provider orders a follow-up test, indication, and a time interval until due. The order can be made or modified immediately after the procedure, when the pathology report is received, or at an office visit. In addition, should guidelines change, procedures can be recalled by indication and the follow-up interval can be altered as necessary. The system then lies dormant until 4 months before the due date for the procedure. At that time the EMR of the patient is queried for a scheduled or completed procedure within 6 months of the due date. If none is found, a notification is sent to the patient s PCP that the patient is due for colonoscopy in the coming months, so that he/she can modify or cancel the order if necessary. If the PCP does not modify the order, 3 months before the procedure due date the patient is sent a standard reminder letter (Appendix 1). One month before the procedure due date, if no procedure has been scheduled or completed the patient is sent an identical reminder letter. Finally, if the due date passes with no procedure scheduled or completed, and no correspondence from the patient, the system alerts administrative staff to call the patient (Figure 1 shows a graphic representation of the follow-up algorithm). Testing of the Automated Follow-up Reminder System The follow-up reminder system is designed to be forward looking (ie, the system is activated at the time of the index procedure. To assess system effectiveness in a practical time period, we evaluated records of all patients who had undergone a colonoscopy at the institution 5 years previously (from August 1, 2004, to February 28, 2005). In this population, we looked for those who were due for a 5-year follow-up study (from August 1, 2009, to February 28, 2010) based on the endoscopist s recom-

3 1168 LEFFLER ET AL GASTROENTEROLOGY Vol. 140, No. 4 Figure 1. Colonoscopy follow-up strategy developed. mendation at the time of the initial study. Records from 2609 patients were evaluated that represented 28 faculty endoscopists and more than 400 referring physicians. Patients were excluded from intervention if they were age 80 or older at the time of the index colonoscopy. All other patients due for colonoscopy during the study time period were included, regardless of indication. Eligible patients were randomized to either the automated follow-up reminder system or standard of care (procedures scheduled by primary care physician or directly by patients, no standardized reminder system used) in a 2:1 ratio using a random number generator. At the conclusion of the study, patients in the control who had not scheduled or completed their colonoscopy were called with a scripted reminder. Data recorded included demographic characteristics, findings on initial colonoscopy, quality of colon preparation, reason for recommended follow-up evaluation, timing of follow-up evaluation, and finding on follow-up evaluation. Patient surveys were administered in the intervention of the study to measure patient satisfaction with the reminder program. A 20-item questionnaire was administered by means of a scripted, telephone interview survey by a qualified interviewer within 2 months of the intervention. The survey was designed specifically for use in this study, but was based on standard, commonly used Likert-formatted items. In addition to patient satisfaction, the survey assessed the perceived clarity of the reminders, reasons for not scheduling a follow-up procedure when applicable, and potential privacy concerns (Appendix 2). The primary end point for the study was the percentage of patients scheduling or undergoing colonoscopy within 6 months of the recommended due date in the intervention vs the standard-of-care cohorts. At the completion of the study, we contacted all patients in the standard-of-care who had not scheduled or completed a colonoscopy to inform them that they were due for a colonoscopy and asked if they were aware of the recommendation for colonoscopy. Statistical Analyses The primary outcome of procedures scheduled or completed in the intervention vs standard-of-care groups was assessed by the Fisher exact test with an a priori significance of All other comparisons were considered exploratory. Relationships between categoric variables were assessed using the chi-square or the Fisher exact tests, whereas the Student t test and the Mann Whitney test were used for ordinal and continuous variables.

4 April 2011 ALERTING SYSTEM 1169 Figure 2. Flow chart of study and major outcomes. *Majority owing to advanced malignancy. **Of those not scheduled: 244 (45.2%) deferred procedure, 13 (2.4%) were not able to be contacted, 13 (2.4%) moved out of system, 3 (0.6%) reported comorbidity, and 1 (0.2%) was deceased. Data were entered into a secure database (Access, Microsoft Office; Microsoft Corp, Redmond, WA) and reviewed for errors before analysis. Statistical analysis was completed using SPSS for Windows (release 17.0, 2008; SPSS, Inc, Chicago, IL). This project was approved by the Beth Israel Deaconess Medical Center Investigational Review Board Protocol #2008-P and funded by Controlled Risk Insurance Company (CRICO)-Risk Management Foundation (Cambridge, MA). The funder played no part in the design or conduct of the trial. This study is registered under clinicaltrials.gov number NCT Results Of 2609 patients identified from August 1, 2004, to February 28, 2005, there were 830 (31.8%) eligible for randomization. The most common reasons for lack of eligibility were that patients were not yet due for follow-up evaluation (32.5%) or had interval colonoscopies performed early for diagnostic purposes (18.7%) (Figure 2). Of the 830 patients, 539 were randomized to the intervention and the remaining 291 were followed up without intervention. Demographics and index colonoscopic factors were similar in the intervention and control cohorts with the exception of a slight predominance of non-english speakers and unspecified race in the intervention (Table 1). The quality of colonoscopy preparations in the index procedures was similar in the control and intervention s. In the intervention, fewer patients had their initial procedures performed for a family history of polyps and more were owing to the presence of symptoms (P.03 and.0001, respectively). Findings, and specifically adenoma detection rates, at the index colonoscopy were similar in the 2 groups. Mean age and sex also were similar between the study groups (Table 1). At the completion of the study, 44.7% of patients randomized to the intervention had procedures scheduled compared with 22.6% in the control group (P.0001). Similarly, at the completion of the study 33.5% of patients in the intervention had procedures

5 1170 LEFFLER ET AL GASTROENTEROLOGY Vol. 140, No. 4 Table 1. Demographics and Index Colonoscopic Factors of the Intervention and Standard-of-Care Cohorts Intervention n 539 (%) Standardof-care n 291 (%) P Female sex 272 (50.4) 146 (50.1).94 Mean age, y Race White 406 (75.3) 236 (81.1).07 All specified non-white 87 (16.1) 39 (13.4).31 African American 49 (9.0) 18 (6.1).18 Non-white, non African 84 (15.6) 37 (12.7).30 American Hispanic or Latino 10 (1.8) 4 (1.3).43 Asian 12 (2.2) 10 (3.4).37 Other 16 (2.9) 7 (2.4).83 Not specified 31 (5.7) 7 (2.4).04 Declined to answer 15 (2.7) 9 (3.0).83 English as primary language 515 (95.5) 287 (98.6).03 Health insurance Private 381 (70.6) 223 (76.6).07 Medicare/Medicaid 143 (26.5) 60 (20.6).06 None 15 (2.7) 8 (2.7) 1.0 Preparation quality.27 Excellent 8 (1.4) 1 (0.3) Good 494 (91.7) 278 (95.5) Fair 29 (5.4) 11 (3.8) Poor 6 (1.1) 1 (0.3) Not reported 2 (0.4) Indication for index colonoscopy a Personal history of polyps 177 (32.8) 85 (29.2).31 Personal history of colon 5 (0.9) 5 (1.7).33 cancer Family history of polyps 20 (3.7) 21 (7.2).03 Family history of colon 143 (26.5) 76 (26.8).93 cancer Colon cancer screening/ 175 (32.4) 109 (37.4).7 surveillance Symptoms b 230 (42.6) 80 (27.4).0001 Inadequate preparation 5 (0.9).17 on prior examination Abnormal radiologic 4 (0.7) 4 (1.3).46 examination No indication noted 1 (0.1) 2 (0.6).28 Findings on prior colonoscopy No significant finding c 408 (75.6) 233 (80.0).29 Adenoma 130 (24.1) 58 (19.9).14 Cancer 1 (0.1) 1.0 a Total percentage is greater than 100% because of multiple indications per patient. b Symptoms include gastrointestinal bleeding, change in bowel habits, and abdominal pain. c Includes hyperplastic polyps, lipomas, uncomplicated diverticulosis, and hemorrhoids. performed compared with 17.8% in the control group (P.0001) (Figure 2 and Table 2). Although the intervention and standard-of-care cohorts overall had similar demographic characteristics, it is notable that non-whites and in particular African Americans were more likely to respond to the follow-up intervention. Although the distribution of races was similar in the 2 cohorts, 43.7% of non-whites had scheduled colonoscopies in the intervention vs only 7.7% in the standard-of-care. This trend was seen across all races, although only whites and African Americans reached statistical significance (Table 2). In patients who underwent colonoscopy, there was an unexpected trend toward more adenomatous polyps detected in the standard-of-care cohort compared with the intervention cohort, 27.9% and 40.8%, respectively, however, this was not statistically significant (Table 3). In the intervention, PCPs canceled or postponed the procedure in 4.4% of cases predominantly owing to a change in patient health status or a disagreement with the recommended interval. The initial letter resulted in the greatest response, with 48.1% of all patients scheduling a colonoscopy doing so after the initial letter. The second letter and the telephone call contributed 17.4% and 9.9%, respectively, whereas 15.4% were scheduled Table 2. Results of Intervention Intervention n 539 (%) Standard-ofcare n 291 (%) Total scheduled 241 (44.7) a 66 (22.6) a.0001 Total completed 181 (33.5) a 52 (17.8) a.0001 Of patients scheduled Female sex 120 (49.7) 34 (51.5).89 Mean age, y Race White 192 (47.3) 60 (25.4).0001 All specified non-white 38 (43.7) 3 (7.7).0001 African American 28 (57.1) 1 (5.6).0002 Non-white, 21/84 (25.0) 5/37 (13.5).23 non African American Hispanic or Latino 3 (30.0) 0.5 Asian 5 (41.7) 1 (10.0).16 Other 2 (12.5) 1 (14.3) 1.0 Not specified 6 (19.4) 0.57 Declined to answer 5 (33.3) 3 (33.3) 1.0 Primary language English 232 (96.2) 65 (98.4).70 Other 9 (3.7) 1 (1.5).70 Health insurance Private 166 (68.8) 50 (75.7).36 Medicare/Medicaid 75 (31.1) 15 (22.7).22 None 1 (1.5).22 Scheduled before 37 (15.3) N/A intervention Scheduled after PCP 22 (9.1) N/A notification Scheduled after first 116 (48.1) N/A letter Scheduled after second 42 (17.4) N/A letter Scheduled after telephone call 24 (9.9) N/A a Percentage is of total number randomized. All other percentages noted are of the number in each cohort with procedures scheduled. P

6 April 2011 ALERTING SYSTEM 1171 Table 3. Findings at Colonoscopy Intervention n 172 (%) a Standard-of-care n 49 (%) a Preparation quality Excellent 1 (0.6) 1.0 Good 153 (88.9) 44 (89.7) 1.0 Fair 14 (8.1) 5 (10.2).77 Poor 3 (1.7) 1.0 Not reported 1 (0.6) 1.0 Indication for procedure b Personal history of polyps 108 (62.7) 34 (69.3).50 Personal history of colon 5 (2.9).59 cancer Family history of polyps 5 (2.9).59 Family history of colon 38 (22.0) 14 (28.5).35 cancer Colon cancer screening/ 35 (20.3) 3 (6.1).02 surveillance Symptoms c 29 (16.8) 10 (20.4).26 Inadequate preparation 1.0 on prior examination Abnormal radiologic 1 (0.6) 1.0 examination No indication noted 2 (1.1) 1.0 Findings No significant finding d 122 (70.9) 28 (57.1).08 Adenoma 48 (27.9) 20 (40.8).11 Cancer 1 (2.0).22 Procedure interrupted 1 (0.6) 1.0 Tissue did not survive processing 1 (0.6) 1.0 a Procedural information was not available on 9 patients in the intervention and 2 patients in the control because the colonoscopy was completed at an outside institution. b Total percentage is greater than 100% because of multiple indications per patient. c Symptoms include gastrointestinal bleeding, change in bowel habits, and abdominal pain. d Includes hyperplastic polyps, lipomas, uncomplicated diverticulosis, and hemorrhoids. before the intervention and 9.1% were scheduled after notification of the PCP. A total of 45.2% of patients deferred colonoscopy without stated reason. In addition, 0.5% of participants reported they had an underlying medical condition that precluded colonoscopy and 0.1% died during the intervention period. A further 2.4% of patients in the intervention contacted investigators to report that they had moved out of system and 2.4% were not able to be reached. In poststudy calls made to individuals in the control, 58% reported being unaware that they were due for colonoscopy. At the end of the study, we also performed telephone patient satisfaction surveys in a random sample of 182 patients from the intervention. Forty-six of 182 (25.2%) attempted calls resulted in completed surveys. In general, patients reported that they found the system helpful, that letters were the preferred method of contact, and many reported not being aware that they P were due for colonoscopy before receiving the reminder. No privacy concerns were noted by patients, and the overall relationship with our institution was improved by the reminder program. The patient satisfaction survey and detailed results of the surveys can be found in Appendix 2. Discussion In this study we created and tested an EMR-based system that reminds patients and providers when follow-up examinations, in this case colonoscopies, are due and provides documentation in the medical record of this communication. Our study is notable for several important results. First, the low-cost intervention almost doubled the rate of recommended examinations during the time period of the study. Second, the intervention was even more effective in minority populations who typically receive care of lower quality, which could lead to improvements in disparities in care for those needing repeat colonoscopies. Finally, the intervention was well received by patients. Very little data exist on colonoscopy follow-up evaluation and most studies in this area focus on clinician adherence to published guidelines, rather than patient adherence to clinician recommendations. A small number of interventions to improve colonoscopy screening adherence have been reported, however, none of these included documentation of individual patient outreach. Further, some interventions that are resource intensive, such as the use of patient navigators, 8,17,18,21 limit adoption and others that use only a standard brochure may be more effective in initial screening than in follow-up evaluation. 19 In this study we found that spontaneous adherence to recommended follow-up evaluation was low. Of the 291 patients in the control of the study, fewer than 25% had scheduled or completed a colonoscopy by the due date. Although it is possible that if we had followed up this group longer that additional patients might have scheduled procedures, it is notable that more than half of patients in the control group called at the end of the study reported being unaware they were due for colonoscopy. Although one could argue that the 33% in the intervention who did have their colonoscopy within the recommended time period is by no means optimal, this percentage represents a significant improvement from standard of care. Although there are some up-front costs associated with the adoption of a follow-up system, once running, well-designed systems can function with little additional burden to the physician or administrative staff. It also should be noted that 100% adherence is not a feasible goal because it is expected that a proportion of patients will either develop interval medical issues that preclude colonoscopy or move their care outside of the institution and become lost to follow-up evaluation. In addition, although there may be concern that adoption

7 1172 LEFFLER ET AL GASTROENTEROLOGY Vol. 140, No. 4 of a follow-up system could increase malpractice risk through system failures both technical and human (incorrect addresses, order entry error), detailed discussion with the risk management group that covers our center and all affiliated groups suggested that these potential risks were substantially less than the risks associated with current practice. Further studies will be necessary to determine whether the documentation provided by this system will be sufficient to reduce the burden of malpractice claims related to inadequate follow-up evaluation. Some practices may be reluctant to devote the resources necessary for the described workflow, however, despite potentially significant up-front costs, once initiated, the cost of maintaining a follow-up reminder system such as this is quite low and may even be revenuegenerating. Further, if follow-up systems were to become standard in new EMR systems, cost would be negligible and benefits would be persistent. It also is notable that the majority of effort was spent on the telephone recall, which was required for only about half of the intervention group. Although we believe that telephone calls are important, this step added only approximately 10% to the overall scheduled procedures, and even without this step the difference in follow-up evaluation between the groups remained highly significant (40.3% vs 22.6%; P.0001). In recent years there has been significant concern regarding overuse of surveillance colonoscopy It is notable that 30.8% of patients studied were found to have adenomatous polyps, a percentage similar to that seen in the previous studies of surveillance colonoscopy, 25 and even the expected polyp detection rate at first screening colonoscopy. 26,27 The high rate of adenoma detection suggests that recommendations for follow-up evaluation were, in general, appropriate. We also would argue that the particular indications and time recommended for follow-up evaluation in this cohort are tangential to the intervention tested. The follow-up system devised generalizes well across multiple different test indications and the potential to improve adherence rates is clear. Further, the intervention was particularly effective in minority populations, which traditionally have lower screening and surveillance rates. 28 This finding suggests that use of similar solutions can reduce the disparities seen in medical care across the US medical system. Although our data provide strong support for the utility of patient-centered reminder systems, we do recognize a number of limitations. First, we focused on patients recommended for 5-year follow-up evaluation, whereas colonoscopy may be recommended at nearly any interval from 6 months to 10 years. We chose this interval for the current study, but we designed our prospective system to accommodate any possible follow-up interval. Five-year follow-up evaluation was chosen to provide homogeneity to the study population at a commonly recommended interval. Second, although patients were randomized, PCPs were not, and most PCPs would have had patients in both the intervention and control s. The reminders provided for the intervention cohort may have increased the PCPs awareness to this matter in general, and hence increased reminders by PCPs to patients in the control of recommended colonoscopies. Such spillover, however, would have biased our results to the null. In addition, despite randomization, more individuals in the study cohort had the index colonoscopy performed for symptoms than in the control group. To investigate whether this factor could have significantly confounded our results, we assessed for a correlation between prior colonoscopy for symptoms and scheduling of a follow-up study. In the control group there was no association (P.128), however, in the study group there was a significant correlation (P.003), with individuals having had the past colonoscopy performed for symptoms being significantly less likely to schedule a follow-up colonoscopy. The reason for this is unclear, but we speculate that individuals who present for colonoscopy with symptoms may, on the whole, be less likely to have another procedure when they are asymptomatic. Regardless of reason, this suggests that if the cohorts had been well matched on reason for index colonoscopy, our finding likely would have been more robust. We also recognize that the patient satisfaction survey was only completed in a small subsample of the study population, and although no concerns were elicited in this group, results may not be fully representative. Finally, we acknowledge that, although successful, this study was conducted at a single center and results may not be generalizable. In summary, our automated, patient-dependent colonoscopy follow-up reminder system significantly improved adherence with recommended surveillance colonoscopy and patient satisfaction. Although this study focused only on colonoscopy, it is expected that this approach would be widely applicable across different procedures and medical specialties. Overall, this study is congruent with the growing body of evidence that EMR systems can improve care and reduce error. 29 At this stage it is unclear how EMR systems will develop and how the general population s growing reliance on electronic means of communication will alter the way patients and clinicians interact. Regardless of whether information is transmitted on paper, in s, texts, or other media, we anticipate that the need for integrated systems to assist in prompting patients to obtain recommended care will increase. As electronic medical record systems are adopted and refined, protocols for notifying and documenting communication regarding recommended follow-up screening and diagnostic procedures should be strongly considered. Supplementary Material Note: To access the supplementary material accompanying this article, visit the online version of

8 April 2011 ALERTING SYSTEM 1173 Gastroenterology at and at doi: /j.gastro References 1. Winawer SJ, Zauber AG, Ho MN, et al. Prevention of colorectal cancer by colonoscopic polypectomy. The National Polyp Study Workgroup. N Engl J Med 1993;329: Citarda F, Tomaselli G, Capocaccia R, et al. Efficacy in standard clinical practice of colonoscopic polypectomy in reducing colorectal cancer incidence. Gut 2001;48: Screening for colorectal cancer: U.S. Preventive Services Task Force recommendation statement. Ann Intern Med 2008;149: Winawer S, Fletcher R, Rex D, et al. Colorectal cancer screening and surveillance: clinical guidelines and rationale update based on new evidence. Gastroenterology 2003;124: Pignone MP, Lewis CL. Using quality improvement techniques to increase colon cancer screening. Am J Med 2009;122: Lloyd SC, Harvey NR, Hebert JR, et al. Racial disparities in colon cancer. Primary care endoscopy as a tool to increase screening rates among minority patients. Cancer 2007;109: Bajracharya SM. An assessment of the perceived barriers and strategies to promoting early detection of colorectal cancer: a practitioners perspective. Int Q Community Health Educ 2006; 26: Nash D, Azeez S, Vlahov D, et al. Evaluation of an intervention to increase screening colonoscopy in an urban public hospital setting. J Urban Health 2006;83: Zimmerman RK, Tabbarah M, Trauth J, et al. Predictors of lower endoscopy use among patients at three inner-city neighborhood health centers. J Urban Health 2006;83: Klabunde CN, Schenck AP, Davis WW. Barriers to colorectal cancer screening among Medicare consumers. Am J Prev Med 2006;30: Denberg TD, Melhado TV, Coombes JM, et al. Predictors of nonadherence to screening colonoscopy. J Gen Intern Med 2005; 20: Colquhoun P, Chen HC, Kim JI, et al. High compliance rates observed for follow-up colonoscopy post polypectomy are achievable outside of clinical trials: efficacy of polypectomy is not reduced by low compliance for follow up. Colorectal Dis 2004;6: Siddiqui AA, Patel A, Huerta S. Determinants of compliance with colonoscopy in patients with adenomatous colon polyps in a veteran population. Aliment Phacol Ther 2006;24: CRICO Risk Managment Foundation. Colorectal Cancer Screening Algorithm. Available: aspx. Date accessed: 3/17/ Braun KL, Fong M, Kaanoi ME, et al. Testing a culturally appropriate, theory-based intervention to improve colorectal cancer screening among Native Hawaiians. Prev Med 2005;40: Myers RE, Turner B, Weinberg D, et al. Impact of a physicianoriented intervention on follow-up in colorectal cancer screening. Prev Med 2004;38: Lebwohl B, Neugut AI, Stavsky E, et al. Effect of a patient navigator program on the volume and quality of colonoscopy. J Clin Gastroenterol October [Epub ahead of print]. 18. Christie J, Itzkowitz S, Lihau-Nkanza I, et al. A randomized controlled trial using patient navigation to increase colonoscopy screening among low-income minorities. J Natl Med Assoc 2008; 100: Denberg TD, Coombes JM, Byers TE, et al. Effect of a mailed brochure on appointment-keeping for screening colonoscopy: a randomized trial. Ann Intern Med 2006;145: Khankari K, Eder M, Osborn CY, et al. Improving colorectal cancer screening among the medically underserved: a pilot study within a federally qualified health center. J Gen Intern Med 2007; 22: Jandorf L, Gutierrez Y, Lopez J, et al. Use of a patient navigator to increase colorectal cancer screening in an urban neighborhood health clinic. J Urban Health 2005;82: Lieberman DA, Holub J, Eisen G, et al. Utilization of colonoscopy in the United States: results from a national consortium. Gastrointest Endosc 2005;62: Schoen RE, Pinsky PF, Weissfeld JL, et al. Utilization of surveillance colonoscopy in community practice. Gastroenterology 2010;138: Laiyemo AO, Pinsky PF, Marcus PM, et al. Utilization and yield of surveillance colonoscopy in the continued follow-up study of the polyp prevention trial. Clin Gastroenterol Hepatol 2009;7: ; quiz, Schoen RE. Surveillance after positive and negative colonoscopy examinations: issues, yields, and use. Am J Gastroenterol 2003; 98: Barclay RL, Vicari JJ, Doughty AS, et al. Colonoscopic withdrawal times and adenoma detection during screening colonoscopy. N Engl J Med 2006;355: Adler A, Pohl H, Papanikolaou IS, et al. A prospective randomised study on narrow-band imaging versus conventional colonoscopy for adenoma detection: does narrow-band imaging induce a learning effect? Gut 2008;57: Fenton JJ, Tancredi DJ, Green P, et al. Persistent racial and ethnic disparities in up-to-date colorectal cancer testing in medicare enrollees. J Am Geriatr Soc 2009;57: Schiff GD, Bates DW. Can electronic clinical documentation help prevent diagnostic errors? N Engl J Med 2010;362: Received November 22, Accepted January 10, Reprint requests Address requests for reprints to: Daniel Leffler, MD, Department of Gastroenterology, Beth Israel Deaconess Medical Center, 330 Brookline Avenue, Boston, Massachusetts dleffler@caregroup.harvard.edu; fax: (617) Acknowledgments The authors greatly appreciate the assistance of Eileen Joyce and Sara O Conner in the Division of Gastroenterology, Beth Israel Deaconess Medical Center, and Kenneth Sands, MD, Department of Medicine, Beth Israel Deaconess Medical Center. Clinical Trials Registration: clinicaltrials.gov #NCT Conflicts of interest The authors disclose no conflicts. Funding Supported by Controlled Risk Insurance Company (CRICO)-Risk Management Foundation. The study sponsor had no role in study design, data collection, analysis, or interpretation.

9 1173.e1 LEFFLER ET AL GASTROENTEROLOGY Vol. 140, No. 4 Appendix 1. Standard Reminder Letter Sent to Patients in the Intervention Cohort Dear Patient, According to our records, you are due for a follow-up colonoscopy in the coming months. Our office would like to remind you to call for your appointment as soon as possible. The phone number for gastroenterology scheduling is After discussion with your primary care physician, if you feel that you do not wish to have a colonoscopy at this time, or if you have scheduled this exam with another provider, please call our office, phone number , to provide us with this information so that we may update our records. Also please check your insurance coverage policies and referral requirements. If a referral is needed, contact your primary care physician s office to request one after setting the date and time with our office. If you have already scheduled the procedure with our office, please disregard this notice. Thank you for your cooperation. Sincerely, Division of Gastroenterology Beth Israel Deaconess Medical Center Appendix 2. Patient Satisfaction Survey Used 1. Were you contacted by BIDMC with a reminder to schedule a colonoscopy procedure within the last few months? Yes/ No/ Not Sure 2. If Yes how were you contacted: a. by 1 letter; b. by 2 letters; c. by 1 letter and a phone call; d. by 2 letters and a phone call; e. by a phone call only. f. not sure/ don t know 3. Did the reminder prompt you to contact your Primary Care Provider? Yes / No 4. Did you schedule an appointment with a gastroenterologist after receiving the reminder? Yes / No 5. If not why not? a. Had already had a repeat colonoscopy elsewhere b. I felt it was too soon c. My PCP felt it was too soon d. Other health conditions prevent me from having this test e. Financial or insurance reasons f. Moved out of area g. Other: 6. If yes would you have scheduled an appointment with a gastroenterologist had you not received the reminder? Definitely would have Definitely would not have 7. Was the message of the letter/letters clear? Very Clear Very Unclear 8. Was the message of the phone call clear? Very Clear Very Unclear 9. Did you find this to be an effective way to communicate information regarding test reminders? Very Effective Not at all Effective 10. What is the best way to contact you in the future regarding test reminders? a. by letters; Best Way Worst Way

10 April 2011 ALERTING SYSTEM 1173.e2 b. by phone; Best Way Worst Way c. by ; Best Way Worst Way d. by text messaging; Best Way Worst Way e. other Best Way Worst Way 11. Would you like to receive reminders regarding colonoscopies and other tests in the future? Definitely Definitely not 12. Did getting this reminder raise any privacy concerns regarding your health information? Definitely Definitely not 13. How did getting a reminder to schedule a colonoscopy affect your relationship with BIDMC? Much Better Relationship Much Worse Relationship 14. Is English your primary language? Yes / No 15. What is the highest grade in school or year of college that you completed? 0. No formal schooling, or did not complete first grade 1. 1st grade 2. 2nd grade 3. 3rd grade 4. 4th grade 5. 5th grade 6. 6th grade 7. 7th grade 8. 8th grade 9. 9th grade th grade th grade th grade year of college years of college or an Associate s degree years of college years of college or a Bachelor s degree or 2 years of graduate school or a Master s degree years or more of graduate school or a Doctoral degree, Law degree 19. Trade, vocational, or technical school 20. GED 16. Are you of Spanish/Hispanic/Latino ethnicity? YES NO 17. What do you consider to be your racial background? (check all that apply) White Black or African American American Indian or Alaskan Native Asian Native Hawaiian or Pacific Islander Other 18. Which of the following best describes your current employment status? (read all options) Employed for wages Self-employed Homemaker

11 1173.e3 LEFFLER ET AL GASTROENTEROLOGY Vol. 140, No. 4 Out of work/ unemployed Retired Unable to work 19. Are there other thoughts you have regarding this medical reminder system or is there information that you would like us to know about you? If so, what is it? Results of Patient Satisfaction Survey Group completing satisfaction survey n 46 (%) 3.95 Sex, female 22 (47.8) Mean age, y 65.2 Underwent colonoscopy 46 (100.0) How were you contacted? Never contacted 3 (6.5) 1 letter 33 (71.7) 2 letters 9 (19.5) 2 letters and a telephone call 1 (2.1) Prompted to contact PCP? 18 (39.1) Prompted to contact gastroenterology? 39 (84.7) Would have scheduled without reminder? a Message of the letter clear? b 1.71 Message of the telephone call clear? b 1.5 Test reminders effective? c 1.37 Best method of contact? c Letter 1.52 Telephone Text messaging 6.29 Would like to receive future reminders? d 1.14 Letters raised privacy concerns? d 7.00 Changed relationship with BIDMC? e 2.42 English primary language? 41 (89.1) Education level: 8th grade 1 (2.1) 10th grade 1 (2.1) 12th grade 3 (6.5) 2 y college 5 (10.8) 4 y college 11 (23.9) 2 y graduate school 1 (2.1) 3 y graduate school 8 (17.3) Master s degree 10 (21.7) Medical degree 2 (4.3) Not indicated 4 (8.6) Racial background White 41 (89.1) Black 2 (4.3) Not indicated 3 (6.5) Employment status Retired 21 (45.6) Employed for wages 16 (34.7) Self-employed 3 (6.5) Out of work/unemployed 2 (4.3) Not indicated 4 (8.6) a Scale 1 7 (mean value); 1 definitely would have, 7 definitely would not have. b Scale 1 7 (mean value); 1 very clear, 7 very unclear. c Scale 1 7 (mean value); 1 very effective, 7 not at all effective. d Scale 1 7 (mean value); 1 definitely, 7 definitely not. e Scale 1 7 (mean value); 1 much better relationship, 7 much worse relationship.