A centrally generated primary care physician audit report does not improve colonoscopy uptake after a positive result on a fecal occult blood test in Ontario ’ s ColonCancerCheck program

Background Timely follow-up of fecal occult blood screening with colonoscopy is essential for achieving colorectal cancer mortality reduction. In the present study, we evaluated the effectiveness of centrally generated, physiciantargeted audit and feedback to improve colonoscopy uptake after a positive fecal occult blood test (fobt) result within Ontario’s population-wide ColonCancerCheck Program. Methods This prospective cohort study used data sets from Ontario’s ColonCancerCheck Program (2008–2011) that were linked to provincial administrative health databases. Cox proportional hazards regression was used to estimate the effect of centralized, physician-targeted audit and feedback on colonoscopy uptake in an Ontario-wide fobt-positive cohort. Results A mailed physician audit and feedback report identifying individuals outstanding for colonoscopy for 3 or more months after a positive fobt result did not increase the likelihood of colonoscopy uptake (hazard ratio: 0.95; 95% confidence interval: 0.79 to 1.13). Duration of positive fobt status was strongly inversely associated with the hazard of follow-up colonoscopy (p for linear trend: <0.001). Conclusions In a large population-wide setting, centralized tracking in the form of physician-targeted mailed audit and feedback reports does not improve colonoscopy uptake for screening participants with a positive fobt result outstanding for 3 or more months. Mailed physician-targeted screening audit and feedback reports alone are unlikely to improve compliance with follow-up colonoscopy in Ontario. Other interventions such as physician audits or automatic referrals, demonstrated to be effective in other jurisdictions, might be warranted.


INTRODUCTION
Fecal occult blood test (fobt) screening depends on followup colonoscopy.As of 2011, only 74.6% of participants with a positive fobt in Ontario's ColonCancerCheck program proceeded to colonoscopy within 6 months.That proportion fell short of early results from European population-wide screening programs and the ColonCancer Check program targets based on Canadian consensus 1 .
To increase follow-up colonoscopy uptake, Colon-CancerCheck introduced patient-and physician-targeted strategies, two of which have already been evaluated 2 .The Screening Activity Report (sar) provides primary care physicians (pcps) with periodic summaries of screening Current Oncology, Vol. 24, No. 1, February 2017 © 2017 Multimed Inc.
participation and identifies fobt-positive patients in their practice who remain outstanding for colonoscopy.In the present study, we evaluate the effectiveness of the sar in improving colonoscopy uptake after a positive fobt result.

METHODS
Our study was approved by the institutional review board at Sunnybrook Health Sciences Centre, Toronto, Ontario.All data were sourced from Ontario health administrative databases, which were linked using unique encoded identifiers and analyzed at the Institute for Clinical Evaluative Sciences 2 .
The population-based prospective cohort comprised all screening-age Ontarians (50-74 years) who received a positive fobt result from 1 September 2008 through 31 December 2010.Participants were excluded if they had a previous diagnosis of colorectal cancer, or if they received a colonoscopy between 31 December 2010 and 1 April 2011.
The main exposure was inclusion in a sar, based on whether the ordering pcp participated in a patient enrolment model (pem) practice.Practitioners participating in the Ontario pem are remunerated using a blended capitation and fee-for-service model that requires patient rostering.In Ontario, pem and traditional models (the latter almost exclusively fee-for-service without stipulation for patient rostering) coexist.The 2011 sars were mailed to pem-participating pcps in February and March of that year, with 97% being mailed during 8-23 March.Physicians practicing in a traditional model did not receive a sar.Only positive fobt results that were outstanding before the end of 2010 were included in the sar.
The primary outcome was time to colonoscopy.Table i presents the covariates by dichotomous sar exposure.Definitions of those variables have been provided elsewhere 2 .
Cox proportional hazards models were used to estimate crude and multivariable-adjusted main effects on the hazard of follow-up colonoscopy within 6 months of 1 April 2011.

RESULTS
Over the 6-month follow-up, 15.1% of the 9661 participants who were fobt-positive for 3 or more months received a follow-up colonoscopy.Approximately half those colonoscopies occurred within the first 2 months.Participants included in a sar (n = 8799) tended to be older and of higher socioeconomic status (Table i).They were also more likely to have higher continuity of care from their ordering pcp.The sar-included participants were moderately more likely to have had a colonoscopy in the preceding 2 years.Conversely, prior colonoscopies among the sar-excluded participants (n = 862) were proportionally more remote.
Crude and adjusted effects indicate that the hazard of colonoscopy was not meaningfully different for sarexposed participants (Table ii).Duration of positive fobt status was observed to be strongly inversely associated w ith hazard of follow-up colonoscopy (p for linear trend: <0.001).

DISCUSSION
Our findings indicate that the first ColonCancerCheck sar did not meaningfully improve colonoscopy uptake for patients who had been fobt-positive for 3 or more months.
Prev ious eva luations of physician-targeted approaches have incorporated multifaceted interventions.In the United States, Veteran Affairs centres and a large Seattle-based health cooperative found that electronic reminders prompting physician feedback [3][4][5] , in addition to other infrastructure upgrades 4 , improved colonoscopy uptake after an abnormal fobt.Another U.S. study found that physician reminders, combined with practice-tailored education, improved the same outcome 6 .Localized U.S. settings have supported the effectiveness of automated referrals 7,8 .Given that the foregoing reports were not examples of population-wide screening, comparisons with Ontario's ColonCancerCheck and European programs might be of limited utility.
The literature evaluating physician-or participanttargeted interventions and aiming specifically to improve colonoscopy uptake after a positive fobt in population-wide settings has been sparse.However, prior evaluations of two ongoing interventions in Ontario's ColonCancerCheck have demonstrated that interventions targeting participants are more effective 2 .
The present evaluation is the first to consider a mailed, physician-targeted intervention that provides practice-specific screening status updates about fobtpositive patients in a population-wide setting.Our findings should be interpreted with a view to two main limitations.First, residual confounding is a possibility.The sar-included participants might have differed from the sar-excluded ones by unmeasured factors beyond inclusion in the sar, either related to having a pem-participating pcp or associated with follow-up colonoscopy.A second limitation is that the intervention effect could be assessed only for participants who had been fobt-positive for 3 or more months.Poorer follow-up with increasing duration of fobt positivity, as supported by our findings (Table ii), suggests that those who do not proceed to colonoscopy directly are less likely to do so, perhaps regardless of intervention type or target.
Although clinical data for ColonCancerCheck screening participants have been found to be 93% accurate, and although pilot evaluations have indicated that Ontario pcps are highly supportive of the sar audit-feedback strategy 9 , refinement of the sar based on pcp feedback is ongoing.In subsequent iterations of the sar, the lag between a positive fobt result and issuance of the pcp report was substantially reduced.Finally, although no formal validation has yet been performed, we have no reason to doubt the accuracy of the sar.

CONCLUSIONS
Mailed physician-targeted audit and feedback reports alone are unlikely to improve compliance with follow-up colonoscopy in Ontario for ColonCancerCheck participants who have remained fobt-positive for 3 or more months.to be effective in localized U.S. settings in addition to participant-targeted correspondence).However, some of these inventions could be onerous, particularly for population-wide screening programs of comparable scope to Ontario's ColonCancerCheck, given that the requisite infrastructure is not already in place.

ACKNOWLEDGMENTS
This study was supported by the Institute for Clinical Evaluative Sciences (ices), which is funded by an annual grant from the Ontario Ministry of Health and Long-Term Care (mohltc).The opinions, results, and conclusions reported in this paper are those of the authors and are independent from the funding sources.Nor endorsement by ices or the Ontario mohltc is intended or should be inferred.Parts of this material are based on data and information compiled and provided by the Canadian Institute for Health Information (cihi) and Cancer Care Ontario (cco).However, the analyses, conclusions, opinions, and statements expressed herein are those of the authors and not necessarily those of cihi.No endorsement by cihi or cco is intended or should be inferred.
We acknowledge Cory Borkhoff for preliminary work in planning and accessing data for this study.

CONFLICT OF INTEREST DISCLOSURES
We have read and understood Current Oncology's policy on disclosing conflicts of interest, and we declare that we have none.

d"
High" indicates that 75% or more of primary care services in the 2 years preceding the index FOBT+ were performed by the same physician.e Each FOBT kit collects 2 samples from each of 3 consecutive spontaneously passed stools, placed on 3 flaps.Of the 2 samples per flap, 1 or more positives define a positive test flap; of all 6 samples, 1 or more positives define a FOBT+ result.f Repeat FOBT within 6 months of the index FOBT+.SES = socioeconomic status; LHIN = local health integration network; ADG = aggregated diagnosis groups.

TABLE I
Distribution of sociodemographic and clinical factors for the cohort of individuals participating in the Ontario-wide ColonCancerCheck program between 1 April 2008 and 31 December 2010 who had a positive fecal occult blood test (FOBT+) and who did not have colonoscopy follow-up before 1 April 2011 Current Oncology, Vol. 24, No. 1, February 2017 © 2017 Multimed Inc.

TABLE I Continued
Ontario health region, ranked in ascending order by colonoscopy rates for 2007 fiscal year.Number of Johns Hopkins ADGs ascertained from physician billing and inpatient hospitalization records during the 12 months preceding the index FOBT+ result.
b c

TABLE II
Crude and multivariable-adjusted hazard ratios for the effects of inclusion in a screening activity report and duration of positive fecal occult blood test (FOBT+) status on hazard of follow-up colonoscopy, 1 April to 1 October 2011 a Adjusted for all covariates listed in TableI.