population based platform to evaluate and monitor the effectiveness of
screening colonoscopy: a population based comparative effectiveness study.
Author: Jaroslaw Regula, MD, PhD
Colorectal cancer (CRC) remains the second leading cause of cancer death in Poland and Europe.1,2 Most CRCs develop from non-malignant precursor lesions called adenomas during long time-interval. This slow development gives an opportunity to intervene with screening tests which may contribute to decreased CRC incidence and mortality by removal of polyps or detect CRC at an early stage. Although colonoscopy is approved by various expert committees as a primary screening tool for the prevention and early detection of CRC for general population, no randomized trials exist to quantify the possible benefit of colonoscopy screening. Three long-term randomized, controlled trials of screening colonoscopy in the general population have recently been initiated but the results will not be available for at least 10 years. Furthermore, the results of these trials may not necessarily be replicated in a public health program. It has been therefore suggested that CRC screening is implemented with experimental design to ensure proper and detailed evaluation of its effects and feasibility as a public health policy.3
The Polish Colonoscopy Screening Platform (PCSP) is a population based comparative effectiveness study designed to evaluate the performance characteristics and effectiveness of screening colonoscopy as a public health policy. In the early phase of implementation following aims are to be covered:
1. The sensitivity of the screening episode and screening program for the detection of CRC expected to become clinically detectable within 2 years.
2. The incidence of advanced CRC in the screening and control group within two years from the date of randomization.
3. A 30-day mortality rate and 8-day hospitalization rate in the screening and control group.
4. The cost-effectiveness of the screening colonoscopy program.
5. Participation in screening colonoscopy in response to various invitation procedures.
6. Participation in screening colonoscopy in response to different bowel preparation regimens offered.
7. The effect of routine video-recording on colonoscopy quality indicators, screenee satisfaction and self assessed pain scores.
Each of the aims listed above will be covered by a separate co-study which together forms the PCSP. Each co-study is going to be run by different leading young investigator.
A long-term follow up of subjects randomly drawn from the population registry to the study allows evaluation of CRC mortality in the screening and control group after 5 years of follow-up. Furthermore, it allows comparing CRC mortality in subjects drawn to receive invitation to screening by the age 55-59 years and 60-64 years (after 10 years of follow-up in each age group; 15 years from starting the study). Subsequent grant applications are going to cover long-term follow-up endpoints.
A national, colonoscopy-based CRC screening program in Poland has been launched in the year 2000.4 Until the end of year 2011 it was run as a non-population based screening (opportunistic) with participants recruited by direct contact with their family or general practitioners. From the year 2012 it started to evolve into a population-based program in which eligible target population is identified through Population Registry and individually invited for screening. Population-based organization of CRC screening maximizes the impact of intervention and ensures high coverage and equity of access.5 It has been agreed to launch the population-based program as experimental public health platform. It means gradual expansion over time with individual-level method of drawing subjects from population registry into screening or control arms among a target population in the implementation phase of the program. Those drawn to be controls will be invited for screening with a 5 years delay (after implementation phase entire target population receives invitation to screening). The design is similar to fecal occult blood testing based public health policy in Finland4 where people are also drawn to FOBT testing or control group. Such design allows public health activities to obtain comparative effectiveness information. This is justified by the fact that it is not possible to provide participation in screening for the whole population in a given time so part of population is provided with delayed service. The costs of screening procedures and simple invitations within the program are financed by the Polish Ministry of Health. However, the Ministry does not cover the costs of scientific evaluation of performance characteristics and effectiveness of the program. This grant application is intended to finance the cutting edge scientific evaluation of the CRC screening program implemented as comparative effectiveness public health platform (PCSP). The concept of evaluating a screening program with experimental design is innovative and gives opportunity to modify or adjust public health policy.
1. Inclusion criteria
All individuals aged 55-64 years living in the counties covered with the organized CRC screening program in the years 2012-2013 are eligible for the study. Eligible individuals will be drawn from the Population Registry (PESEL Registry) and assigned in a 1:1 ratio to the group invited to screening colonoscopy in the year of drawing or in 5 years time (this group will not be informed about their status as controls over 5 years until the time of invitation to screening). Drawing will be stratified by sex and age cohorts. Figure 1 shows simplified scheme of this process. Over 2 years time approximately 240 000 eligible individuals living in 45 counties spread in different voivoedships will be included in the program and evaluation.
2. Exclusion criteria
Following individuals will be excluded:
· Message from neighbor/family/post office on death of screenee (not updated in Population Registry)
· Resident abroad (not updated in Population Registry)
· Return of unopened letter of invitation and/or reminder (address unknown)
· Diagnosis of CRC before the date of draw
3. Exclusion criteria for screening colonoscopy (these individuals will not be offered any screening but will be included in the intention-to-screen analysis).
· Individuals in need of long-lasting attention and nursing services (somatic or psychosocial, mental retardation)
· Severe cardiac or lung disease limiting routine daily activity
· Proctocolectomy or colectomy
Excluded individuals with suspicious symptoms will be advised to consult their physician for further investigations.
A coordination center of the Polish Colorectal Cancer Screening Program mails a letter of invitation with the appointment date and hour to every individual drawn to the screening group. The control group members are not informed of their status as controls. Letter of invitation accompanied with an informational folder and reply form will be mailed 6-7 weeks prior to the appointment. Respective local screening center mails a reminder letter to non-responders 3 weeks prior to the appointment. Local screening centre manages invitee’s responses by phone or mail, confirms appointment dates and sends bowel preparation in case of acceptance to attend. The colonoscopy procedure and the bowel preparation are offered free of charge for the invitees.
Split-dose 4L polyethylene glycol (PEG) regimen with a one-day low-fiber diet will be used as a standard in the study. Alternative bowel preparation regimens will be compared with the standard one as a part of the PCSP (see brief summaries of co-studies planned within the PCSP).
Standard video colonoscopes will be used to perform examinations. At colonoscopy, all detected CRC precursor lesions are removed whenever feasible, and other pathological findings are biopsied. The choice to immediately remove more difficult lesions or postpone removal until pre-prepared therapeutic colonoscopy is evaluated by considering the safety of individuals. Insertion of the endoscope is discontinued if the subject expresses discomfort or a wish to stop.
Screening episode refers to both screening and evaluation test results needed to complete one screening round.
Clinically detectable CRC
CRC that would be detected by a diagnostic evaluation of symptomatic patient.
Episode sensitivity for CRCs expected to become clinically detectable
The estimated sensitivity of the episode for CRCs expected to become clinically detectable (with the reference to those screened) within some time interval
Program sensitivity for CRCs expected to become clinically detectable
The estimated sensitivity of the entire program for CRCs expected to be clinically detectable (with the reference to the invited population) within some time interval.
Adenoma detection rate
The proportion of screened subjects in whom at least one adenomatous lesion is identified.
Cecal intubation rate
The passage of the colonoscope tip to a point proximal to the ileocecal valve and visualization of the entire cecum.
Statistical methods and sample size:
In each of co-study different statistical methods are applied; these are briefly described in the appropriate section (see below). The PCSP population of 240 000 people includes 120 000 drawn to the screening group with estimated 30 000 (25%) attending colonoscopy and 120 000 controls. Large PCSP population provides adequate sample size for each of the planned co-studies. For each co-study a P value of less than 0.05 is considered to indicate statistical significance.
Brief summary of co-studies within the Polish Colonoscopy Screening Platform:
1. The sensitivity of the screening episode and screening program for the detection of CRC expected to become clinically detectable within 2 years (task for a young doctor).
Screening test targets precancerous or early cancerous lesion that are expected to become symptomatic in the future. However, no direct observation of the lesion in the preclinical phase is available because they are removed in a consequence of screening. Therefore, the sensitivity of screening for the detection of CRC expected to become clinically detectable within certain time has to be done indirectly, by measuring failure of screening (interval CRCs). The entire PCSP population of 240 000 people will be included in this study and followed-up with Cancer Registries to capture the incidence of CRC. The sensitivity of the screening episode/program will be estimated using the incidence method.3
2. The incidence of advanced CRC in the screening and control group within two years from the date of draw (task for a PhD student).
The effect of screening is evaluated by comparing CRC mortality between those invited for screening and controls. Screening colonoscopy should decrease the incidence of advanced CRC in the screening group prior to reduction in CRC mortality and can therefore be used as an early indicator of its effectiveness. The aim of this study is to compare the incidence of advanced CRC in the screening and control group within two years from the date of draw in the intention to screen analysis. The entire PCSP population of 240 000 people will be included in this study and followed-up with Cancer Registries to capture the incidence of advanced CRC (regional or distant metastases according to Cancer Registry format).
3. A 30-day mortality rate and 8-day hospitalization rate in the screening and control group (task for a young doctor).
Although European Union guidelines recommend capturing 30-day mortality rate and 8-day hospitalization rate, it has never been done in organized colonoscopy screening.5 The aim of this study is to compare 30-day mortality rate and 8-day hospitalization rate in the screening and control groups of the PCSP. The entire PCSP population of 240 000 people will be included in this study. The control group will have randomly assigned virtual date of invitation to screening and virtual date of screening colonoscopy. A 30-day mortality rate and 8-day hospitalization rate following factual or virtual invitation to screening and factual or virtual screening colonoscopy will be captured through PESEL Registry and National Health Fund Registry, respectively.
4. The cost-effectiveness of the screening colonoscopy program (task for a young doctor).
The estimated cost-effectiveness of the Polish CRC Screening Program is unknown. The aim of this study is to assess the cost-effectiveness of the PCSP using computer models based on a Markov process. The input data (screening uptake rates, prevalence of CRC and adenomas, costs etc.) will be directly taken from the PCSP, whenever possible.
5. Participation in screening colonoscopy in response to various invitation procedures (task for a PhD student and a medical student).
Participation in screening colonoscopy remains unsatisfactory. An advanced notification letter has been shown to increase participation in fecal occult blood testing programs. The aim of this study is to evaluate the effect of advance notification letter on participation in primary screening colonoscopy. 8,000 men and women aged 55 to 64 years will be assigned in a 1:1 ratio to receive advance notification letter followed by standard invitation in 2 weeks time and reminding letter 3 weeks later or standard invitation and reminding letter 3 weeks later. The sample size calculated was 6,800 to detect 3% difference in participation rate (25.0% vs. 28.0%) with 80% power.
6. Participation in screening colonoscopy in response to different bowel preparation regimens offered (task for a PhD student and medical student).
Bowel preparation for colonoscopy is unacceptable for some screening candidates and can therefore have detrimental effect on compliance. In a limited number of studies magnesium citrate plus sodium picosulphate was shown to have comparable cleansing efficacy but better tolerability than PEG based solutions. The aim of this study is to evaluate the effect of magnesium citrate plus sodium picosulphate on participation in screening colonoscopy. 16,000 men and women aged 55 to 64 years will be randomly assigned in a 1:1 ratio to a group offered split dose magnesium citrate plus sodium picosulphate or split dose 4L PEG solution in case of positive response to screening invitation. The sample size calculated was 15,100 to detect 2% difference in participation rate (25.0% vs. 27.0%) with 80% power. Secondary endpoints will be adequacy of bowel preparation, adenoma detection rate and cecal intubation rate at screening colonoscopy.
7. The effect of routine video-recording on colonoscopy quality indicators, screene’ satisfaction and self assessed pain scores (task for a young doctor and a medical student).
Although, video recording of colonoscopy has been proposed as a means to enhance the quality of colonoscopy, its’ effectiveness is uncertain. The aim of this study is to evaluate the impact of routine video recording of screening colonoscopies on adenoma detection rate, cecal intubation rate or a screenee satisfaction. This is a cluster randomized controlled trial in which eight screening centers will be assigned either to routine video recording of all screening colonoscopies or no video recording (current standard). In the first year cecal intubation rate and adenoma detection rate of all participating endoscopists will be monitored through dedicated common database and screenee satisfaction using Gastronet questionnaire.6 In the second year in four randomly chosen screening centers routine video recording will be implemented. The primary endpoint is difference in adenoma detection rate between the first and second year of the study. The sample size calculated was 4991 per group to detect 3.5% difference between 5% (from 25% to 30%) and 1.5% (from 25% to 26.5%) in adenoma detection rate improvement with 5% level of significance between video recording and not video recording centers with 80% power.
The role of international partners:
Geir Hoff, Micheal Bretthauer, Mette Kalager (Centre for Colorectal Cancer Screening, The Cancer Registry of Norway, Oslo University Hospital, Oslo, Norway), Nea Malila (Mass Screening Registry, Finnish Cancer Registry, Helsinki, Finland)
Listed partners are world experts in colorectal cancer screening. Contacts and links between these partners and the Department of Gastroenterology, Medical Centre for Postgraduate Education and Institute of Oncology in Warsaw exist already for many years. The listed group has designed the first randomized study on colonoscopy for colorectal cancer known as NordICC study. The listed people are responsible in their countries for implementation of national colorectal cancer screening programs. Their experience and know-how is of great value and assurance that quality of planned activities in Poland will be the highest possible. The list of tasks for the international group is as follows:
1. Know-how on implementation of CRC screening as an experimental public health policy
2. Provision of the Gastronet system (patient oriented evaluation of the quality of screening examinations) to evaluate screenee satisfaction
3. Joint design and evaluation of the results of some co-studies
4. Participation in the international group meetings to discuss research cooperation and scientific network organization.
1. Ferlay J, Parkin DM, Steliarova-Foucher E. Estimates of cancer incidence and mortality in Europe in 2008. Eur J Cancer 2010;46: 765-81.
2. Wojciechowska U, Didkowska J, Zatonski W. Cancer in Poland in 2008. National Cancer Registry of Poland, the Maria Sklodowska-Curie Memorial Cancer Center and Institute of Oncology 2010; Warsaw, Poland.
3. Malila N, Oivanen T, Malminiemi O, Hakama M. Test, episode, and programme sensitivities of screening for colorectal cancer as a public health policy in Finland: experimental design. BMJ 2008;337:a2261. doi: 10.1136/bmj.a2261.
4. Regula J, Rupinski M, Kraszewska E, et al. Colonoscopy in colorectal-cancer screening for detection of advanced neoplasia. N Engl J Med 2006;355: 1863-72.
5. European guidelines for quality assurance in colorectal cancer screening and diagnosis – first edition. Segnan N, Patnick J, von Karsa L (eds.). European Commission, Publications Office of the European Union, Luxembourg. 2010.
6. Hoff G, Bretthauer M, Huppertz-Haus G, et al. The Norwegian Gastronet project: Continous quality improvement of colonoscopy in 14 Norwegian centres. Scand J Gastroenterol 2006;41: 481-7.
Figure 1. Simplified scheme of the drawing process.