Evaluating the impact of public health initiatives on trends in fecal occult blood test participation in Ontario

Dear Author, Here are the final proofs of your article. Please check the proofs carefully. All communications with regard to the proof should be sent to [email protected]. Please note that at this stage you should only be checking for errors introduced during the production process. Please pay particular attention to the following when checking the proof: - Author names. Check that each author name is spelled correctly, and that names appear in the

correct order of first name followed by family name. This will ensure that the names will be indexed correctly (for example if the author’s name is ‘Jane Patel’, she will be cited as ‘Patel, J.’).

- Affiliations. Check that all authors are cited with the correct affiliations, that the author who will receive correspondence has been identified with an asterisk (*), and that all equal contributors have been identified with a dagger sign (†).

- Ensure that the main text is complete. - Check that figures, tables and their legends are included and in the correct order. - Look to see that queries that were raised during copy-editing or typesetting have been resolved. - Confirm that all web links are correct and working. - Ensure that special characters and equations are displaying correctly. - Check that additional or supplementary files can be opened and are correct. Changes in scientific content cannot be made at this stage unless the request has already been approved. This includes changes to title or authorship, new results, or corrected values. How to return your corrections Returning your corrections via online submission: - Please provide details of your corrections in the online correction form. Always indicate the line

number to which the correction refers. Returning your corrections via email: - Annotate the proof PDF with your corrections. - Send it as an email attachment to: [email protected]. - Remember to include the journal title, manuscript number, and your name when sending your

response via email. Note: in order to ensure timely publication, if we do not hear from you within 48 hours we may take the decision to sign-off the article on your behalf and proceed to publication. After you have submitted your corrections, you will receive email notification from our production team that your article has been published in the final version. All changes at this stage are final. We will not be able to make any further changes after publication. Kind regards, BioMed Central Production Team 2

1 RESEARCH ARTICLE Open Access

2 Evaluating the impact of public health initiatives3 on trends in fecal occult blood test participation4 in Ontario5 Gladys N Honein-AbouHaidar1, Linda Rabeneck2,3,4,5,6, Lawrence F Paszat6,7, Rinku Sutradhar2,6,6 Jill Tinmouth4,5,6,7,8 and Nancy N Baxter5,6,9*789101112131415

16 Abstract

17 Background: Since the publication of two randomized controlled trials (RCT) in 1996 demonstrating the18 effectiveness of fecal occult blood test (FOBT) in reducing colorectal cancer (CRC) mortality, several public health19 initiatives have been introduced in Ontario to promote FOBT participation. We examined the effect of these20 initiatives on FOBT participation and evaluated temporal trends in participation between 1994 and 2012.

21 Method: Using administrative databases, we identified 18 annual cohorts of individuals age 50 to 74 years eligible22 for CRC screening and identified those who received FOBT in each quarter of a year. We used negative binomial23 segmented regression to examine the effect of initiatives on trends and Joinpoint regression to evaluate temporal24 trends in FOBT participation.

25 Results: Quarterly FOBT participation increased from 6.5 per 1000 in quarter 1 to 41.6 per 1000 in quarter 7226 (January-March 2012). Segmented regression indicated increases following the publication of the RCTs in 199627 (Δ slope = 6%, 95% CI = 4.3-7.9), the primary care physician financial incentives announcement in 2005 (Δ slope = 2.2%,28 95% CI = 0.68-3.7), the launch of the ColonCancerCheck (CCC) Program (Δ intercept = 35.4%, 95% CI = 18.3 -54.9),29 and the CCC Program 2-year anniversary (Δ slope = 7.2%, 95% CI = 3.9 – 10.5). Joinpoint validated these findings and30 identified the specific points when changes occurred.

31 Conclusion: Although observed increases in FOBT participation cannot be definitively attributed to the various32 initiatives, the results of the two statistical approaches suggest a causal association between the observed increases in33 FOBT participation and most of these initiatives.

34Keywords: Public health policy, Colorectal cancer screening, Epidemiologic study

35 Background36 The population health burden of colorectal cancer (CRC)37 in Canada is substantial [1]. In Ontario, Canada, CRC is38 the second cause of cancer mortality [1]. Screening for39 CRC can reduce the burden of this disease. Three land-40 mark randomized controlled trials (RCTs) published be-41 tween 1993 and 1996 demonstrated that biennial use42 of the fecal occult blood test (FOBT), coupled with colonos-43 copy in those who test positive, resulted in a 15% reduction

44in CRC mortality [2-4]. The publication of these RCTs moti-45vated policy makers to make various efforts to promote46FOBT participation in Ontario.47In February 2001, the Canadian Task Force on Preventive48Health Care (CTFPHC) published guidelines recommend-49ing FOBT as a CRC screening test for average risk indi-50viduals aged 50 to 74 years (Level A Recommendation)51[5]. The dissemination of these guidelines into clinical52practice was passive and without any mechanism to pro-53mote adherence.54In July 2005, the Ministry of Health and Long-Term55Care (MOHLTC) of Ontario announced new financial in-56centives for CRC screening targeting primary care physi-57cians (PCPs) in patient enrolment model (PEM) types of

* Correspondence: [email protected] for Health Policy Management and Evaluation, University ofToronto, Toronto, ON, Canada6Institute for Clinical Evaluative Sciences, Toronto, ON, CanadaFull list of author information is available at the end of the article

© 2014 Honein-AbouHaidar et al.; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms ofthe Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use,distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons PublicDomain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in thisarticle, unless otherwise stated.

Honein-AbouHaidar et al. BMC Cancer 2014, 14:537http://www.biomedcentral.com/1471-2407/14/537

58 practice (50% of Ontario physicians at that time) [6]. Eli-59 gible PCPs received end of fiscal year bonuses based on60 the proportion of enrolled patients who received FOBT61 prior to March 31st of each year. The bonus amount in-62 creases as the proportion of screened patients increases,63 e.g. if 20% of enrolled patients are screened, the PCP re-64 ceives $440; if 50% are screened, the PCP receives $2,200.65 The first bonus submission was on April 1st of 2006 for66 FOBT screening of enrolled patients from April 1 200567 through March 31 2006 [7].68 In April 2008, Cancer Care Ontario, Ontario’s provin-69 cial cancer agency responsible for cancer services, and70 the MOHLTC launched the ColonCancerCheck (CCC)71 Program, the first province-wide organized CRC screen-72 ing program in Canada. The CCC Program recommends73 FOBT every 2 years for average risk individuals age 50 to74 74 years and colonoscopy for those who test positive [8].75 An intense but temporary public media campaign and a76 PCP educational program marked the launch of the CCC77 Program. Starting from fiscal year 2008, PCPs became eli-78 gible to receive up to $4,000 if 70% of their enrolled pa-79 tients were screened [9-12].80 April 2010 marked the CCC Program 2- year anniver-81 sary. In addition to ongoing PCP screening practices, the82 CCC Program rolled out recall and reminder letter inter-83 ventions. Recall letters were sent out to those who were84 FOBT negative in the first round of screening inviting85 them to be re-screened. These recall letters were sent in86 August 2010 for those who completed FOBT in the pre-87 vious 24–30 months and in December 2010, a reminder88 letter was sent for those who had not yet undergone89 FOBT screening [13].90 The goal of this population-based time trend study was91 to examine the effect of the publication of the RCTs and92 the CTFPHC guidelines, the announcement of PCP finan-93 cial incentives, the launch of the CCC Program, and the94 programmatic correspondence following the CCC Program95 2-year anniversary on FOBT participation in Ontario and96 to evaluate temporal trends in FOBT participation between97 April 1st 1994 and March 31st 2012.

98 Methods99 The Research Ethics Board of St. Michael’s Hospital in100 Toronto approved this study.

101 Data sources102 We used four data holdings including the Registered103 Persons Database (RPDB), the Ontario Health Insurance104 Plan (OHIP) database, the Ontario Cancer Registry (OCR),105 and the Canadian Institute for Health Information106 Discharge Abstract Database (CIHI-DAD). These data107 holdings are housed at the Institute for Clinical Evaluative108 Sciences (ICES) [14]. Each data record collected at109 ICES comes with personal identifier, usually a health

110card number. Using a secure ICES algorithm, each health111card number is assigned a unique encrypted ICES number112(IKN). Once records in a data set have an IKN assigned,113the identifying information is stripped off the file and114the data become de-identified. Researchers have access115to the de-identified data only. The unique IKN is used to116link the various data sets.117The RPDB is a roster of all permanent residents and118refugees eligible for coverage under the Ontario Health119Insurance Plan, which contains demographic information120including an individual’s date of birth, sex, date of death121(where applicable), and changes in eligibility for health in-122surance coverage. The OHIP database contains informa-123tion about all claims for physician and laboratory services124provided to Ontario residents since July 1991. The OCR is125a registry of all Ontario residents diagnosed with cancer126since 1964. The OCR captures over 95% of cancer cases in127Ontario [15]. The CIHI-DAD contains information from128hospitalization records, abstracted since April 1988.

129Study cohorts130All persons eligible for OHIP aged 50 to 74 years were131identified from the RPDB at the beginning of each fiscal132year from 1994 to 2012. Using IKN, we linked these co-133horts to OCR and CIHI-DAD to exclude individuals diag-134nosed with CRC or Inflammatory Bowel Disease before135April 1st of each year to approximate cohorts of individ-136uals at average risk for CRC. (Additional file 1: Diagnostic137and OHIP procedure codes).138We used OHIP database to identify those who re-139ceived CRC screening tests in each fiscal year and in the140previous ten years (Additional file 1). For persons with141multiple claims in a fiscal year, we included the first ser-142vice date for FOBT; for persons with multiple claims in143the previous 10 years we included the most recent ser-144vice date for this time period.145The data were analyzed by quarter of a fiscal year. For146each quarter, we included all individuals due for CRC147screening in our denominator; individuals who under-148went FOBT during the quarter formed our numerator.149We applied the following exclusions to approximate a150population that was due for CRC screening:

1511- At the beginning of each quarter, we excluded those152who died in the previous quarter(s) of the same year;1532- At the end of each quarter, we excluded those154who were up-to-date with CRC screening as defined155as having: FOBT within two years; a flexible156sigmoidoscopy or barium enema within five years;157or a colonoscopy within ten years.

158Statistical analysis159We used two statistical methods. We used a segmented160regression analysis to compare changes in trends in FOBT

Honein-AbouHaidar et al. BMC Cancer 2014, 14:537 Page 2 of 7http://www.biomedcentral.com/1471-2407/14/537

161 participation before and after initiatives including: publica-162 tion of RCTs (1996), publication of the CTFPHC guide-163 lines (2001), announcement of PCP financial incentives164 (2005), launch of the CCC Program (2008), and the pro-165 grammatic correspondence following the CCC Program166 2-year anniversary (2010). In this analysis, a dummy167 variable (INT) coded 0 before and 1 after the ex-168 pected time of each intervention, and an interaction169 term (INT*Timeafter) were added to the model as sug-170 gested by Wagner et al. [16]. The dummy variable (INT)171 indicates change in intercept, the interaction term indi-172 cates change in slope (Detailed procedure of statistical173 analysis is shown in Additional file 2). A change in slope174 or intercept was considered statistically significant if the175 95% confidence interval did not include zero. Data were176 analyzed using SAS software 9.3. [17].177 Because segmented regression uses pre-defined points,178 the results may mask the specific date when the actual179 change in trend occurred [18]. We, therefore, conducted180 a Joinpoint regression (ver. 4.0) a technique that enables181 trend modeling without pre-defined points [19,20]. We182 fitted the joinpoint regression model as follows: we183 used FOBT count in each quarter as the numerator,184 individuals due for CRC screening (denominator) as an185 “offset term”, and the quarter as the regressor variable.186 We estimated the quarterly percent change (QPC),187 i.e. rate of change in slope between joinpoints, the inter-188 cept of each joinpoint, and corresponding 95% confidence189 intervals using the following parameters: 1) Grid Search190 method; 2) Bayesian Information Criteria model selection

191method; 3) up to 6 joinpoints for each model; 4) a mini-192mum of 5 quarters between two joinpoints; and 5) Poisson193variance [21]. The trend was considered statistically sig-194nificant if the 95% confidence interval of the QPC did not195include zero [18,20-24].

196Results197Cohort characteristics198From fiscal year 1994 to 2012, there were 72 quarters. In199each quarter, we identified 198,000 to 207,000 individuals200due for CRC screening. Quarterly FOBT participation in-201creased from 6.5 in quarter 1 to 41.6 per 1000 in quarter20272 with a peak in quarter 69 (April-June, 2011), after the203programmatic correspondence of the CCC Program (45.9204per 1000). Figure F11 demonstrates an overall increase in205FOBT participation between 1994 and 2012 that was not206uniform throughout the time period. Participation slowly207increased between 1996 and 2005; more rapid increases208occurred after 2005.

209Segmented regression results210We plotted the observed and adjusted quarterly rates of211FOBT participation in each quarter (Figure 1). The re-212sults of the segmented regression analysis are shown in213Table T11.214There was a statistically significant increase in slope in215FOBT participation following the publication of the216RCTs in 1996 (change in slope = 6.1%, 95% CI = 4.3-7.9),217and the announcement of PCP financial incentives (change218in slope = 2.2%, 95% CI = 0.7-3.8). The launch of the CCC

Figure 1 Observed rates and segmented regression adjusted rates of fecal occult blood test (FOBT) participation per 1000, Ontario,1994–2012. Observed rate = (FOBT completed per quarter/ population due for CRC screening per quarter)* 1000. Adjusted rate = (Exp(log rate-offset))*1000. Rates are connected by a binomial regression line. Dashed vertical lines indicate quarter when the following initiatives were enacted: RCT: Publication of the second and third randomized controlled trials in November 1996. CTFPHC: Publicationof the Canadian Task Force on Preventive Health Care guidelines for CRC screening in February 2001. Announcement of PCP financialincentives in July 2005. CCC Program launch, April 2008. CCC Program 2-year anniversary, April 2010. The regression model was expressed as:Log FOBT completed per quarter=population due for CRC screening per quarterð Þ ¼α þ β1quarter þ

X5

j¼1

βj INT quarter ≥ INTj� � þ

X5

j¼1

βj INT� Timeafter quarter ≥ INTjð Þ� quarter−INTjð Þð Þ:

Honein-AbouHaidar et al. BMC Cancer 2014, 14:537 Page 3 of 7http://www.biomedcentral.com/1471-2407/14/537

219 Program was associated with increase in intercept (change220 in intercept = 35.4%, 95% CI = 18.3-54.9) followed by a221 decrease in slope (change in slope =−9.75%, 95% CI = −12-222 7.4). An increase in slope was detected following the CCC223 Program correspondence in 2010 (change in slope: 7.2%,224 95% CI = 3.9-10.5). Other changes in intercept and slope225 were not statistically significant (Table 1).

226 Joinpoint results227 We plotted the observed rates of FOBT participation per228 quarter and the Joinpoint location in FigureF2 2. The re-229 sults of the Joinpoint regression analysis are shown in230 TableT2 2.231 Joinpoint regression analysis identified five joinpoints232 and six distinct segments. The change in slopes between

233Joinpoints and those from the segmented regression234analysis converged. Joinpoint regression identified the235specific point in time when change occurred, the slope236between joinpoints, and the intercept at each joinpoint.237An increase in slope started two quarters following RCT238publication in 1996 (QPC = 3.8%, 95% CI = 3.4-4.2),239followed by another increase in slope starting from240the quarter PCP financial incentives were announced241(QPC = 7.4%, 95% CI = 6.4-8.5). There was an immedi-242ate increase in intercept following the CCC Program243launch (Intercept = 62.1, 95% CI: 59-64.9), a decrease244in slope three quarters after the launch (QPC = −5.5%,24595% CI = −9.9-0.9), and an increase in slope one quarter246before the CCC Program 2-year anniversary (QPC = 8.2%,24795% CI = 0.9-16) (Table 2).

t1:1 Table 1 Segmented regression analysis showing changes in intercept and changes in slope on FOBT participation ratest1:2 following each initiative, 1994-2012

t1:3 Initiative (Segment) Intercept* Slope ^

t1:4 Change in Intercept (Δ) 95% CI Change in slope (Δ) 95% CI

t1:5 Baseline (April 94-October 96) −2.4¥‡ (−3.9–0.9)

t1:6 RCT (October 96-January 01) −3.2 (−13.1-7.9) 6.1‡ (4.3-7.9)

t1:7 CTFPHC (January 01-July 05) 6.8 (−2.8-17.3) 0.3 (−0.6–1.3)

t1:8 FI (July 05- April 08) −1.5 (−11.9–10.1) 2.2‡ (0.7– 3.8)

t1:9 CCC launch (April 08-April 10) 35.4‡ (18.3–54.9) −9.7‡ (−12–7.4)

t1:10 2-year anniversary (April 10 –March 12) 13.5 (−1.6–30.9) 7.2‡ (3.9–10.5)

t1:11 *Difference between pre and post initiative intercepts interpreted as step change and calculated as QPC = (exp βINTi -1 )* 100.t1:12 ^Difference between pre and post initiative slopes taking into account the trend before the initiative and calculated as QPC = (exp βINT*TIMEi -1 )* 100.t1:13 ¥Baseline slope.t1:14 ‡Statistically significant if 95% confidence interval does not cross zero.t1:15 RCT: Publication of the second and third randomized controlled trials in November 1996.t1:16 CTFPHC: Publication of the Canadian Task Force on Preventive Health Care guidelines for CRC screening in February 2001.t1:17 FI: Announcement of PCP financial incentives in July 2005.t1:18 CCC launch: ColonCancer Check program (CCC) Program launch, April 2008.t1:19 2 - year anniversary: ColonCancerCheck Program 2-year anniversary, April 2010.

Figure 2 Observed rates of FOBT participation per 1000 and joinpoint location determined by Joinpoint regression analysis, Ontario,1994–2012. Observed rate = (FOBT completed per quarter/population due for CRC screening per quarter)* 1000. * joinpoint location. Dashedvertical lines indicate quarter when the following initiatives were enacted: RCT: Publication of the second and third randomized controlled trialsin November 1996. CTFPHC: Publication of the Canadian Task Force on Preventive Health Care guidelines for CRC screening in February 2001.Announcement of PCP financial incentives in July 2005. CCC Program launch, April 2008. CCC Program 2-year anniversary, April 2010.

Honein-AbouHaidar et al. BMC Cancer 2014, 14:537 Page 4 of 7http://www.biomedcentral.com/1471-2407/14/537

248 Discussion249 Since 1994, FOBT participation has increased substantially250 in Ontario. We observed an overall increase in quarterly251 participation from 6.5 per 1000 in April 1994 to 41.6 per252 1000 in March, 2012. Participation slowly increased be-253 tween 1994 and 2005 followed by a more rapid increase254 between 2005 and 2012. Although we cannot definitively255 attribute the observed increases in FOBT participation to256 the initiatives made to promote participation, the con-257 vergence of the two statistical approaches suggest a causal258 association between the observed increases in FOBT par-259 ticipation and the publication of the RCTs, introduction of260 PCP financial incentives and CCC Program launch and261 programmatic correspondence, but not publication of the262 CTFPHC guideline.263 We previously reported the results of a segmented re-264 gression to investigate the effect of the launch of the CCC265 Program on FOBT participation in Ontario over a 6 year266 time period (2005 to 2011) [25]. Our current study im-267 proves upon this analysis by evaluating 18 years of data268 allowing examination of initiatives before CCC Program269 launch, enabling the evaluation of CCR program in context270 of previous trends in FOBT uptake, and evaluation of the271 programmatic correspondence of the 2-year anniversary272 of the CCC Program. Further, this study uses two dif-273 ferent statistical approaches, each with specific advantages.274 Segmented regression analysis allowed us to estimate the275 changes in intercepts and slopes following each intervention276 accounting for baselines trends, a robust method for meas-277 uring the effect of an intervention when randomization or278 identification of a control group are impractical [16,26-28].279 Joinpoint analysis enabled identification of specific points280 in time when changes occurred, and provided estimates of281 the actual intercept and slope for each segment.282 Previously, we reported a significant increase in FOBT283 participation (change in intercept) immediately following284 the launch of the CCC Program; we attributed the in-285 crease to the public media campaign [25]. This increase

286was followed by a downtrend at the end of the screening287period, a concern for policy makers (Dr. Linda Rabeneck,288personal communication, January 2009). In the current289study, we found that this downtrend was reversible and290was observed again after the CCC Program 2-year anni-291versary, i.e. a peak after the programmatic correspondence292followed by a drop at the end of the study period. Fluctu-293ation in trends following the introduction of public pol-294icies are reported in the literature [29]. In this instance,295however, a periodic trend in FOBT participation with a296peak every 2 years has likely been introduced, in keeping297with the date of program launch and program recommen-298dation of biennial FOBT screening. Future studies need to299examine if this biennial periodicity will persist and the im-300pact on endoscopic and surgical resources.301In 1996, results of RCTs demonstrated that screening302with FOBT reduces CRC mortality and in 2001 the303CTFPHC strongly endorsed CRC screening with FOBT.304Given this evidence, why increases in FOBT participation305before 2005 were modest? Integration of evidence into306clinical practice has always been challenging [30-32]. Davis307et al. indicated that in order for guidelines to be translated308into practice, there must be intervention strategies to309reinforce their adoption such as reminder systems and aca-310demic detailing [31]. Passive strategies, including mailing311or publication of guidelines, have little impact on adoption312[31]. Because there was no mechanism to actively promote313the CTFPHC guidelines, the modest increase in the use of314FOBTafter their publication is not surprising.315We demonstrated a marked change in participation316following the introduction of financial incentives and the317programmatic correspondence after the CCC Program3182-year anniversary, indicating these initiatives were likely319the reasons for the rapid increase in participation after3202005. In terms of financial incentives, studies show mixed321effects on performance varying between no effect at all322[33,34] and improved performance [35-38]. Certain factors323have proven to be effective in improving performance.

t2:1 Table 2 Joinpoint regression analysis for FOBT participation in Ontario, 1994–2012 showing actual intercept at eacht2:2 joinpoint and actual slope between joinpoints

t2:3 Identified segment Join-point Intercept Slope

t2:4 Intercept£ 95% CI QPC^ 95% CI

t2:5 April 1994-April 1997 0.3 (0.15-0.47) −1.7¥ (−3.9–0.5)

t2:6 April 1997-July 2005 Q 13 0.15‡ (0.02-0.28) 3.8‡ (3.4–4.2)

t2:7 July 2005-October 2009 Q 46 0.03 (−0.47-0.53) 7.4‡ (6.4–8.5)

t2:8 October 2009-January 2010 Q 59 62.1‡ (59–64.9) −5.5‡ (−9.9–0.9)

t2:9 January 2010-January 2011 Q 64 0.01 (−4.5-4.5) 8.2‡ (0.9-16)

t2:10 January 2011-March 2012 Q 68 6‡ (3.1-8.9) −1.4 (−5.5–2.9)

t2:11 £Intercept at each joinpoint calculated as (exp β INTERCEPT i).t2:12 ^Quarterly percent change (slope) between joinpoints calculated as (exp β SLOPEi -1)* 100.t2:13 ¥Baseline trend.t2:14 ‡Statistically significant if 95% confidence interval does not cross zero.

Honein-AbouHaidar et al. BMC Cancer 2014, 14:537 Page 5 of 7http://www.biomedcentral.com/1471-2407/14/537

324 Custers et al. indicate that financial incentives that take325 into account the size of the bonus, and baseline perform-326 ance often succeed in improving performance [39]. In this327 study, two factors may explain the improved performance.328 First, the size of the reward may have motivated some329 physicians to change their screening routines [34,40]. Sec-330 ond, when baseline performance is relatively modest, the331 introduction of bonuses is more likely to have an impact332 [41]. Our findings that participation increased following333 the programmatic correspondence are consistent with334 those from previous studies that suggested that reminder335 letters were associated with increased screening participa-336 tion [42-46].337 Our study has limitations. In observational studies, it is338 difficult to infer a causal association between an interven-339 tion and observed trends [47]. We are examining changes340 in FOBT participation occurring in a complex health sys-341 tem, and factors other than those evaluated in this study342 may have contributed to changes in trend. However, seg-343 mented regression analysis controls for secular trends,344 i.e. reasons other than the effect of initiatives, by introdu-345 cing a term in the model to test the effect of the interven-346 tion over and above the secular trend [16,48].

347 Conclusion348 FOBT participation in Ontario slowly increased between349 1994 and 2005 followed by a more rapid increase be-350 tween 2005 and 2012. The results of the two statistical351 methods suggest a causal association between those in-352 creases and publication of the RCTs, introduction of PCP353 financial incentives and CCC Program launch and pro-354 grammatic correspondence, but not the CTFPHC guideline355 publication. We particularly observed a marked increase356 after the introduction of the CCC Program in 2008. Al-357 though this increase cannot be solely attributed to the CCC358 Program, evidence from the literature suggests that orga-359 nized screening programs are effective in increasing partici-360 pation. Furthermore, we noted a marked increase following361 the programmatic correspondence after the CCC Program362 2-year anniversary. With the information available, it is rea-363 sonable to conclude that the marked increase in participa-364 tion since 2008 might well reflect the impact of the CCC365 Program on FOBT participation.

366 Additional files367369 Additional file 1: Diagnostic and Ontario Health Insurance (OHIP)370 procedure codes. *International Classification of Diseases, 9th and 10th

371 revisions, Clinical Modification.

372 Additional file 2: Detailed procedure of statistical analysis.

373 Abbreviations374 CRC: Colorectal cancer; RCTs: Randomized controlled trials; FOBT: Fecal occult375 blood test; CTFPHC: Canadian Task Force on Preventive Health Care;376 MOHLTC: Ministry of Health and Long-Term Care; PCPs: Primary care

377physicians; PEM: Patient enrolment model; CCC: ColonCancerCheck;378RPDB: Registered persons database; OHIP: Ontario Health Insurance Plan;379OCR: Ontario Cancer Registry; CIHI-DAD: Canadian Institute for Health380Information Discharge Abstract Database; ICES: Institute for Clinical Evaluative381Sciences; IKN: Encrypted ICES number.

382Competing interests383The authors declare that they have no competing interests.

384Authors’ contributions385NB and GHA conceived, designed, drafted, and revised the manuscript.386LR, JT, LP conceived, designed, and revised the manuscript critically for387intellectual content. GHA conducted the analysis. RK interpreted the results388and revised the manuscript critically for intellectual content. All authors read389and approved the final manuscript.

390Acknowledgments391This research was supported through a Cancer Care Ontario research grant392and Canadian Cancer Society Research Institute award. Dr Baxter holds the393Cancer Care Ontario Health Services Research Chair.

394Author details3951Division of Support, System and Outcomes, University Health Network,396Toronto, ON, Canada. 2Dalla Lana School of Public Health, University of397Toronto, Toronto, ON, Canada. 3Prevention and Cancer Control, Cancer Care398Ontario, Toronto, ON, Canada. 4Department of Medicine, University of399Toronto, Toronto, ON, Canada. 5Institute for Health Policy Management and400Evaluation, University of Toronto, Toronto, ON, Canada. 6Institute for Clinical401Evaluative Sciences, Toronto, ON, Canada. 7Sunnybrook Research Institute,402Toronto, ON, Canada. 8ColonCancerCheck Program, Cancer Care Ontario,403Toronto, ON, Canada. 9Department of Surgery and Li Ka Shing Knowledge404Institute, St. Michael’s Hospital, Toronto, ON, Canada.

405Received: 26 February 2014 Accepted: 9 July 2014406Published: 25 July 2014

407References1. 408Canadian Cancer Society’s Advisory Committee on Cancer Statistics:

409Canadian Cancer Statistics 2013. Toronto, On: Canadian Cancer Society; 2013.2. 410Mandel JS, Bond JH, Church TR, Snover DC, Bradley GM, Schuman LM,

411Ederer F: Reducing mortality from colorectal cancer by screening for412fecal occult blood. Minnesota Colon Cancer Control Study. N Engl J Med4131993, 328(19):1365–1371.

3. 414Kronborg O, Fenger C, Olsen J, Jorgensen OD, Sondergaard O: Randomised415study of screening for colorectal cancer with faecal- occult-blood test.416Lancet 1996, 348(9040):1467–1471. doi:10.1016/S0140-6736(96)03430-7.

4. 417Hardcastle JD, Chamberlain JO, Robinson MH, Moss SM, Amar SS,418Balfour TW, James PD, Mangham CM: Randomised controlled trial of419faecal-occult- blood screening for colorectal cancer. Lancet 1996,420348(9040):1472–1477.

5. 421CTFPH: Canadian task force on preventive health care. 2001, [February4222010]; Available from: Q1www.canadiantaskforce.ca.

6. 423Kantarevic J, Kralj B, Weinkauf D: Enhanced fee-for-service model and424physician productivity: evidence from Family Health Groups in Ontario.425J Health Econ 2011, 30(1):99–111 [Comparative Study].

7. 426Li J, Hurley J, Decicca P, Buckley G: Physician Response to Pay- for-Performance:427Evidence from a Natural Experiment. Health Econ 2013, 23(8):962–978.

8. 428Colon Cancer Check 2008 Program Report. Ontario: Toronto Cancer Care;4292010. [September 11, 2013]; Available from: Q2www.cancercare.on.ca.

9. 430Ministry of Health and Long Term Care Teams Preventive Care Bonus-Tracking431and Exclusion Codes. 2006. [cited February 2013]; Available from:432 Q3http://www.anl.com/pages/mohguide.htm.

10. 433Colon Cancer Check Fecal Occult Blood Testing (FOBT) (Bulletin 4471).434[Bulletin]. 2008. [February 2013]; Available from: Q4http://www.health.gov.on.435ca/en/pro/programs/coloncancercheck/role.aspx.

11. 436Colon Cancer Check Fecal Occult Blood Testing (FOBT), (Bulletin 4482)437[Bulletin]. 2008. [February 2013]; Available from: http://www.health.gov.on.438ca/en/pro/programs/coloncancercheck/role.aspx.

12. 439Partnering to Successfully Launch Ontario’s Colorectal Screening Program.440Cancer Care Ontario: Colon Cancer Check regional primary care441engagement report; 2009.

Honein-AbouHaidar et al. BMC Cancer 2014, 14:537 Page 6 of 7http://www.biomedcentral.com/1471-2407/14/537

13.442 ColonCancerCheck. [Accessed September 2011 Accessed September 2011];443 Available atQ5 http://health.gov.on.ca/en/ms/coloncancercheck/.

14.444 Institute for Clinical Evaluative Sciences. Toronto: 2013. Available from:Q6 445 http://www.ices.on.ca/index.html.

15.446 Robles SC, Marrett LD, Clarke EA, Risch HA: An application of capture-recapture447 methods to the estimation of completeness of cancer registration. J Clin448 Epidemiol 1988, 41(5):495–501.

16.449 Wagner AK, Soumerai SB, Zhang F, Ross-Degnan D: Segmented regression450 analysis of interrupted time series studies in medication use research.451 J Clin Pharm Ther 2002, 27(4):299–309 [Research Support, Non-U.S. Gov’t452 Research Support, U.S. Gov’t, P.H.S.].

17.453 The Data Analysis for This Paper was Generated Using SAS/STAT Software.454 Version 9.3. Cary, NC, USA: SAS Institute Inc; 2013.

18.455 Gagne M, Robitaille Y, Hamel D, St-Laurent D: Firearms regulation and456 declining rates of male suicide in Quebec. Inj Prev 2010, 16(4):247–253.

19.457 Kim HJ, Fay MP, Feuer EJ, Midthune DN: Permutation tests for joinpoint458 regression with applications to cancer rates. Stat Med 2000, 19(3):335–351.

20.459 National. Cancer Institute. Joinpoint Regression Program, Version 4.0 [Internet].460 Bethesda, MD: National Cancer Institute; 2013. [cited 2013 June 10];461 Available from:Q7 http://surveillance.cancer.gov/joinpoint/.

21.462 Zhichang JZ, Zhenguo Q, Hatcher J: Joinpoint Trend Analysis of Cancer463 Incidence and Mortality Using Alberta Data Alberta Canadian Partnership464 Against Cancer. Available from:Q8 http://www.cancerview.ca/idc/groups/public/465 documents/webcontent/csen_cproj_fy0910q3_joinpoint.pdf.

22.466 Lee YC, Huang YT, Tsai YW, Huang SM, Kuo KN, McKee M, Nolte E: The467 impact of universal National Health Insurance on population health:468 the experience of Taiwan. BMC Health Serv Res 2010, 10:225 [Evaluation469 Studies].

23.470 Edwards BK, Ward E, Kohler BA, Eheman C, Zauber AG, Anderson RN,471 Jemal A, Schymura MJ, Lansdorp-Vogelaar I, Seeff LC, van Ballegooijen472 M, Goede SL, Ries LA: Annual report to the nation on the status of473 cancer, 1975–2006, featuring colorectal cancer trends and impact of474 interventions (risk factors, screening, and treatment) to reduce future475 rates. Cancer 2010, 116(3):544–573.

24.476 Marrett LD: User Documentation for Surveillance Analytic Software: JoinPoint:477 Cancer Care Ontario. 2009.

25.478 Honein-Abouhaidar GN, Baxter NN, Moineddin R, Urbach DR, Rabeneck L,479 Bierman AS: Trends and inequities in colorectal cancer screening480 participation in Ontario, Canada, 2005–2011. Cancer Epidemiol 2013,481 37(6):946–956.

26.482 Lagarde M: How to do (or not to do) … Assessing the impact of a policy483 change with routine longitudinal data. Health Policy Plan 2012, 27(1):76–83484 [Research Support, Non-U.S. Gov’t].

27.485 Shardell M, Harris AD, El-Kamary SS, Furuno JP, Miller RR, Perencevich EN:486 Statistical analysis and application of quasi experiments to antimicrobial487 resistance intervention studies. Clin Infect Dis 2007, 45(7):901–907488 [Research Support, N.I.H., Extramural Research Support, U.S. Gov’t, Non-P.H.S.489 Research Support, U.S. Gov’t, P.H.S.].

28.490 Zarychanski R, Dennis J, Singh H: Challenges of population-based colorectal491 cancer screening and the importance of time-trend analysis when evaluating492 system change. Cancer Epidemiol 2013, 37(6):957–958.

29.493 Wilson N, Thomson G, Grigg M, Afzal R: New smoke-free environments494 legislation stimulates calls to a national Quitline. Tob Control 2005,495 14(4):287–288 [Letter].

30.496 Anderson LM, May DS: Has the use of cervical, breast, and colorectal497 cancer screening increased in the United States? Am J Public Health 1995,498 85(6):840–842.

31.499 Davis DA, Taylor-Vaisey A: Translating guidelines into practice. A systematic500 review of theoretic concepts, practical experience and research501 evidence in the adoption of clinical practice guidelines. CMAJ 1997,502 157(4):408–416 [Review].

32.503 Grimshaw JM, Russell IT: Effect of clinical guidelines on medical practice: a504 systematic review of rigorous evaluations. Lancet 1993, 342(8883):1317–1322505 [Research Support, Non-U.S. Gov’t].

33.506 Hillman AL, Ripley K, Goldfarb N, Nuamah I, Weiner J, Lusk E: Physician507 financial incentives and feedback: failure to increase cancer screening in508 Medicaid managed care. Am J Public Health 1998, 88(11):1699–1701.

34.509 Grady KE, Lemkau JP, Lee NR, Caddell C: Enhancing mammography510 referral in primary care. Prev Med 1997, 26(6):791–800.

35.511 Rosenthal MB, Frank RG, Li Z, Epstein AM: Early experience with pay-for-512 performance: from concept to practice. JAMA 2005, 294(14):1788–1793.

36. 513Doran T, Fullwood C, Gravelle H, Reeves D, Kontopantelis E, Hiroeh U,514Roland M: Pay-for-performance programs in family practices in the515United Kingdom. N Engl J Med 2006, 355(4):375–384.

37. 516Grossbart SR: What’s the return? Assessing the effect of “pay- for-517performance” initiatives on the quality of care delivery. Med Care Res Rev5182006, 63(1 suppl):29S–48S.

38. 519Kouides RW, Bennett NM, Lewis B, Cappuccio JD, Barker WH, LaForce FM:520Performance-based physician reimbursement and influenza521immunization rates in the elderly. Am J Prev Med 1998, 14(2):89–95.

39. 522Custers T, Hurley J, Klazinga NS, Brown AD: Selecting effective incentive523structures in health care: A decision framework to support health care524purchasers in finding the right incentives to drive performance.525BMC Health Serv Res 2008, 8(1):66.

40. 526Town R, Kane R, Johnson P, Butler M: Economic incentives and physicians’527delivery of preventive care: a systematic review. Am J Prev Med 2005,52828(2):234–240.

41. 529Sabatino SA, Habarta N, Baron RC, Coates RJ, Rimer BK, Kerner J, Coughlin530SS, Kalra GP, Chattopadhyay S: Interventions to increase recommendation531and delivery of screening for breast, cervical, and colorectal cancers by532healthcare providers systematic reviews of provider assessment and533feedback and provider incentives. Am J Prev Med 2008, 35(1 Suppl):S67–S74534[Research Support, U.S. Gov’t, Non-P.H.S.Research Support, U.S. Gov’t, P.H.S.535Review].

42. 536Lee JK, Reis V, Liu S, Conn L, Groessl EJ, Ganiats TG, Ho SB: Improving fecal537occult blood testing compliance using a mailed educational reminder.538J Gen Intern Med 2009, 24(11):1192–1197 [Comparative Study Randomized539Controlled Trial Research Support, Non-U.S. Gov’t].

43. 540Eaker S, Adami HO, Granath F, Wilander E, Sparen P: A large population-based541randomized controlled trial to increase attendance at screening for cervical542cancer. Cancer Epidemiol Biomarkers Prev 2004, 13(3):346–354 [Clinical Trial543Multicenter Study Randomized Controlled Trial Research Support, Non-U.S.544Gov’t].

44. 545Morrell S, Taylor R, Zeckendorf S, Niciak A, Wain G, Ross J: How much does546a reminder letter increase cervical screening among under-screened547women in NSW? Aust N Z J Public Health 2005, 29(1):78–84 [Clinical Trial548Randomized Controlled Trial].

45. 549King ES, Rimer BK, Seay J, Balshem A, Engstrom PF: Promoting mammography550use through progressive interventions: is it effective? Am J Public Health5511994, 84(1):104–106 [Clinical Trial Comparative Study Randomized Controlled552Trial Research Support, U.S. Gov’t, P.H.S.].

46. 553Page A, Morrell S, Chiu C, Taylor R, Tewson R: Recruitment to mammography554screening: a randomised trial and meta-analysis of invitation letters and555telephone calls. Aust N Z J Public Health 2006, 30(2):111–118 [Comparative556Study Evaluation Studies Meta-Analysis Randomized Controlled Trial].

47. 557Shadish WR, Cook TD, Campbell DT: Experimental and Quasi- Experimental558Designs for Generalized Causal Inference. 2002.

48. 559Ramsay CR, Matowe L, Grilli R, Grimshaw JM, Thomas RE: Interrupted time560series designs in health technology assessment: lessons from two561systematic reviews of behavior change strategies. Int J Technol Assess562Health Care 2003, 19(04):613–623.

563doi:10.1186/1471-2407-14-537564Cite this article as: Honein-AbouHaidar et al.: Evaluating the impact of565public health initiatives on trends in fecal occult blood test participation566in Ontario. BMC Cancer 2014 14:537.

Submit your next manuscript to BioMed Centraland take full advantage of:

• Convenient online submission

• Thorough peer review

• No space constraints or color figure charges

• Immediate publication on acceptance

• Inclusion in PubMed, CAS, Scopus and Google Scholar

• Research which is freely available for redistribution

Submit your manuscript at www.biomedcentral.com/submit

Honein-AbouHaidar et al. BMC Cancer 2014, 14:537 Page 7 of 7http://www.biomedcentral.com/1471-2407/14/537

Author Query Form

Journal: BMC Cancer

Title: Evaluating the impact of public health initiatives on trends in fecal occult blood testparticipation in Ontario

Authors: Gladys N Honein-AbouHaidar, Linda Rabeneck, Lawrence F Paszat, Rinku -Sutradhar, Jill Tinmouth, Nancy N Baxter

Article: 1471-2407-14-537

Dear Authors,

During production of your paper, the following queries arose. Please respond to these by annotating your proofs withthe necessary changes/additions. If you intend to annotate your proof electronically, please refer to the E-annotationguidelines. We recommend that you provide additional clarification of answers to queries by entering your answers onthe query sheet, in addition to the text mark-up.

Query No. Query Remark

Q1 URL: Please check that the following URLs are working. If not, please providealternatives: www.canadiantaskforce.ca

Q2 URL: Please check that the following URLs are working. If not, please providealternatives: www.cancercare.on.ca

Q3 URL: Please check that the following URLs are working. If not, please providealternatives: http://www.anl.com/pages/mohguide.htm

Q4 URL: Please check that the following URLs are working. If not, please providealternatives: http://www.health.gov.on.ca/en/pro/programs/coloncancercheck/role.aspx

Q5 URL: Please check that the following URLs are working. If not, please providealternatives: http://health.gov.on.ca/en/ms/coloncancercheck/

Q6 URL: Please check that the following URLs are working. If not, please providealternatives: http://www.ices.on.ca/index.html

Q7 URL: Please check that the following URLs are working. If not, please providealternatives: http://surveillance.cancer.gov/joinpoint/

Q8 URL: Please check that the following URLs are working. If not, please providealternatives: http://www.cancerview.ca/idc/groups/public/documents/webcontent/csen_cproj_fy0910q3_joinpoint.pdf