Costs and Cost Effectiveness of a Health Care Provider–Directed Intervention to Promote Colorectal Cancer Screening Among Veterans
http://www.100md.com
《临床肿瘤学》
the Institute for Healthcare Studies, Department of Medicine, Northwestern University
Robert H. Lurie Comprehensive Cancer Center and Divisions of General Internal Medicine, Hematology/Oncology, Geriatric Medicine, and Gastroenterology, Department of Medicine, Northwestern University Feinberg School of Medicine
Disease Management Association of America
Veterans Affairs Mid-West Center for Health Services and Policy Research, Jesse Brown Veterans Affairs Medical Center/Lakeside Community-Based Outpatient Clinic
Center for Pharmacoeconomic Research, College of Pharmacy, University of Illinois at Chicago, Chicago, IL.
ABSTRACT
PURPOSE: Colorectal cancer screening is underused, particularly in the Veterans Affairs (VA) population. In a randomized controlled trial, a health care provider–directed intervention that offered quarterly feedback to physicians on their patients' colorectal cancer screening rates led to a 9% increase in colorectal cancer screening rates among veterans. The objective of this secondary analysis was to assess the cost effectiveness of the colorectal cancer screening promotion intervention.
METHODS: Providers in the intervention arm attended an educational workshop on colorectal cancer screening and received confidential feedback on individual and group-specific colorectal cancer screening rates. The primary end point was completion of colorectal cancer screening tests. Sensitivity analyses investigated cost-effectiveness estimates varying the data collection methods, costs of labor and technology, and the effectiveness of the intervention.
RESULTS: Rates of colorectal cancer screening for the intervention versus control arms were 41.3% v 32.4%, respectively (P < .05). The incremental cost-effectiveness ratio was $978 per additional veteran screened based on feedback reports generated from manual review of records. However, if feedback reports could be generated from information technology systems, sensitivity analyses indicate that the cost-effectiveness estimate would decrease to $196 per additional veteran screened.
CONCLUSION: An intervention based on quarterly feedback reports to physicians improved colorectal cancer screening rates at a VA medical center. This intervention would be cost effective if relevant data could be generated by existing information technology systems. Our findings may have broad applicability because a 2005 Medicare initiative will provide the VA electronic medical record system as a free benefit to all US physicians.
INTRODUCTION
The importance of screening for colorectal cancer as a component of evidence-based medicine is well documented.1 Colorectal cancer screening is recommended for average-risk individuals who are 50 years old and older; screening methods include an annual fecal occult blood test (FOBT), a flexible sigmoidoscopy or double-contrast barium enema every 5 years, and a colonoscopy every 10 years.1 However, national screening rates for colorectal cancer are less than 50% and much lower than rates reported for prostate, cervical, and breast cancers.2-5 Individuals who are racial or ethnic minorities and/or who are of lower socioeconomic status have especially low rates of colorectal cancer screening, which may contribute in part to the disproportionately large impact of colorectal cancer on those population subgroups.6,7 This issue is especially relevant in the Veterans Affairs (VA) health care system, the largest integrated delivery system in the country, which provides care for 2% of the US population.8 In 2003, colorectal cancer screening rates were the lowest of the 17 routinely reported performance measures included in the External Peer Review Program of VA medical centers.9
Recently, we reported that a health care provider–directed intervention resulted in a 9% increase in colorectal cancer screening rates at a large urban VA medical center.10 For this intervention to be adopted by others, information is needed on the effectiveness, costs, and cost effectiveness of the intervention. Previous studies have not evaluated the economics of screening promotion efforts for persons who are at average risk for the development of colorectal cancer.11 Studies of individuals at average risk for breast and cervical cancers indicate considerable variability in the cost of health promotion interventions, ranging from less than $10 to more than $2,000 per additional woman screened for each of these two cancers (Table 1). The objective of this study is to evaluate the costs and the cost effectiveness of our health care provider–directed intervention for colorectal cancer among veterans.
METHODS
Provider-Directed Continuous Quality Improvement Intervention to Promote Colorectal Cancer Screening Among Veterans
The Chicago VA Health Care System General Medicine Program operates as three independent firms. In the recent intervention, one firm served as the intervention arm, and all of the physicians in this firm attended an introductory session outlining the importance of colorectal cancer screening as well as tips for efficient communication to veterans, many of whom have poor literacy skills. The intervention was based on a continuous quality improvement framework.23
Approximately every quarter, the providers were invited to attend feedback sessions, during which they received information on the firm's colorectal cancer screening recommendation rate and patient adherence to recommended tests. The providers also received confidential information on their individual recommendation and adherence rates. These sessions included reviews of colorectal cancer screening guidelines and practical strategies to communicate with patients with low literacy skills in busy, high-volume VA primary care clinics. The providers also received quarterly feedback reports on their own colorectal cancer screening recommendation and performance rates and comparative information on the firm-specific rates of these measures.
In the control firm, physicians did not attend the introductory session or receive quarterly feedback reports. Physicians in both firms received the current standard of an electronic, patient-specific reminder through the electronic medical record for colorectal cancer screening eligibility. A research assistant abstracted information on colorectal cancer screening recommendation and performance from the electronic medical records of veterans in the general medicine clinic who were 50 years of age or older, were of average risk for colorectal cancer, and had not returned a three-card FOBT within the previous year or undergone a screening flexible sigmoidoscopy/colonoscopy procedure within the previous 5 years. This information was entered into a relational database. The study statistician generated provider-specific and global firm-specific reports that contained colorectal cancer screening recommendation and performance rates for feedback to health care providers in the intervention arm. Sixty and 53 health care providers cared for 1,015 and 963 eligible study patients for the intervention and controlfirms, respectively.10
Flow Diagram
The flow diagram (Fig 1) represents the sequence of events from the initial random assignment of the firms. This analysis allowed us to identify the relevant clinical outcomes in the intervention and control groups (colorectal cancer screening recommendation and performance) and the resources used in generating the quarterly feedback reports used in the intervention arm.
Assumptions for the Economic Model
We made the following two simplifying assumptions: (1) health care provider participation in the educational/feedback sessions occurred at no cost to the provider because the sessions were conducted over lunchtime and (2) provider-patient discussions about colorectal cancer screening were conducted at no cost because it was opportunistic and required minimal time investment by both parties. These assumptions are consistent with other published cost-effectiveness studies.15-22,24-27
Costs of Obtaining Provider-Specific Colorectal Cancer Screening Rates via Manual Review of Electronic Medical Records
We identified and determined all inputs used for the intervention based on generation of feedback reports by manual review of electronic medical records. The research team (three researchers and two statisticians) provided time estimates for the following: manual review of electronic medical records to identify eligible patients and to assess whether screening had been recommended and completed; data input and analysis for report generation; and development and implementation of the promotional program itself. Costs were derived for the entire 25-month study period. The specific resources included personnel, overhead (amount assigned by institution for normal operating expenses; eg, electricity, which is assumed to be 26% based on VA cost-accounting estimates), and costs related to conducting the feedback sessions. Personnel cost estimates were derived from actual salary and benefit data. A base salary of approximately $30,000 per year for a research assistant was assumed, with a fringe benefit rate of 23.5%. Health care resources not related to implementing the intervention (eg, screening) were excluded (Table 2).
Costs of Generating Provider-Specific Colorectal Cancer Screening Rates via Information Technology Systems
In a second analysis, the data included in the feedback reports for the intervention arm were assumed to be generated with VA information technology support rather than by manual review of electronic medical records. Given the ease of electronic data acquisition within the VA healthcare system, we viewed this to be a logical next step for intervention implementation. Furthermore, a new government program was recently announced that will improve the access of physician practices to electronic medical record technology, increasing the likelihood that feedback could be automated at a larger proportion of non-VA sites in the near future.28 Expert opinion from an information technology specialist, two health economists, and two health services researchers resulted in an estimate of a 90% reduction in personnel time used in the manual review of patients' electronic medical records.
Data Analysis
The cost effectiveness of this provider-directed intervention was evaluated. The incremental cost-effectiveness ratio was derived based on the difference in the costs assigned to the intervention and control groups divided by the difference in colorectal cancer screening rates between the intervention and control groups. Estimates were derived for the incremental cost per additional veteran screened for colorectal cancer (cost effectiveness) and for the average cost of the colorectal cancer screening promotion program per provider involved (mean costs of the intervention per provider). Sensitivity analyses investigated cost-effectiveness estimates varying the source of patient data (derived with information technology review of the electronic medical record v abstracted from the medical record by a research assistant), labor costs, and the effectiveness of the intervention. Because health care policy and funding decisions may involve cost-effectiveness data, the economic analysis was conducted from the third party–payer perspective.12-14 For contextual purposes, the incremental cost-effectiveness ratios (measured as dollars per additional veteran screened) for colorectal cancer screening promotion derived in this intervention were compared with ratios previously reported for other cancer screening promotional efforts (Table 3).
RESULTS
Colorectal cancer screening was recommended for 76.0% of patients in the intervention firm and for 69.4% of controls (P = .02). Rates of colorectal screening (effectiveness data) for the intervention versus control arms were 41.3% v 32.4%, respectively (P = .003).10
Obtaining Provider-Specific Colorectal Cancer Screening Rates via Manual Review of Electronic Medical Records
The total annual cost of the health care provider–directed intervention, using manual review of electronic medical records to generate provider-specific colorectal cancer screening rate data, was $46,741 (Table 2). Over the 25-month intervention period, the total cost of the colorectal cancer screening promotional effort was $86,753, and the mean cost per individual provider was $1,446. The incremental cost-effectiveness ratio for the intervention using manual review of electronic medical records was $978 per additional veteran screened for colorectal cancer.
Sensitivity estimates based on 95% CIs around the effectiveness estimate (38.2% to 52.6%) indicated that the estimated cost per additional veteran screened in the intervention arm ranged from $767 to $3,213. Additional sensitivity analysis estimates based on a 10% variation around the point estimate for costs of manual review of medical records ranged from $78,078 to $95,428 for the 25-month period, and associated incremental cost-effectiveness estimates ranged from $877 to $1,072 per additional veteran screened.
Generating Provider-Specific Colorectal Cancer Screening Rates via Information Technology Systems
We also evaluated the costs of the promotional program assuming that the health care provider–specific colorectal cancer screening rates could be generated directly from information technology supported databases. In this scenario, data collection costs for generating provider-specific feedback rates were estimated to decrease by 90% to $7,702. The overall costs of using information technology–generated feedback data would be $17,431, and the mean costs per provider would be $291. The incremental cost-effectiveness ratio was estimated as $196 per additional person screened for colorectal cancer with feedback data generated by information technology systems.
Corresponding sensitivity estimates for the information technology–supported colorectal cancer screening intervention based on 95% CIs around the point estimate for the effectiveness of the intervention ranged from $156 to $655 per additional veteran screened. Additional sensitivity analysis estimates for the information technology–supported intervention based on a 10% variation around the point estimate for the costs of electronically generating colorectal cancer screening rates ranged from $15,688 to $19,174 for the 25-month period, corresponding to incremental cost-effectiveness estimates of $176 to $215 per additional veteran screened.
DISCUSSION
Our economic analysis provides the first estimates of the costs and cost effectiveness of colorectal cancer screening promotion among veterans at average risk of developing colorectal cancer who receive care at an urban VA medical center. An innovative health care provider–directed intervention, consisting primarily of quarterly feedback sessions highlighting colorectal cancer screening rates of individual providers and average rates of all providers in a particular clinic, resulted in a 9% absolute increase in screening rates over a 2-year period.10 The cost of the feedback program, when manual review of electronic medical records was used to obtain screening rates, was $978 per additional veteran screened for colorectal cancer. However, if the screening data could be generated directly from information technology–supported data systems, the cost could be dramatically reduced to $196 per additional veteran screened.
Our findings can be interpreted in the context of key elements for evaluating the cost effectiveness of cancer screening promotional studies suggested by Andersen et al11 and outlined in Table 3. These concepts address the following seven factors: (1) basing interventions around screening tests that are known to be cost effective; (2) targeting average-risk individuals who have not previously participated in screening efforts; (3) evaluating quality-of-life concerns of participants involved in cancer screening programs; (4) focusing on interventions that are likely to be cost effective; (5) formally assessing the cost effectiveness of the promotion efforts during a randomized intervention; (6) recognizing that the unique characteristics of the patient population, physicians, and health care system play an important role in the promotional efforts; and (7) addressing the sustainability of the promotional program. This health care provider–directed intervention addresses six of the seven proposed lessons, with the exception of including prospective quality-of-life assessments and cost-effectiveness data in terms of quality-adjusted life years. However, the purpose of this analysis was not to elucidate the cost effectiveness of the screening technology (which has been repeatedly investigated)29 but rather to study the promotion strategy itself and to provide further insight into how to improve the cost effectiveness of future colorectal cancer screening promotion strategies.11
Our economic assessment highlights the importance of considering data collection costs when evaluating the feasibility of providing quarterly colorectal cancer screening rate information to health care providers. If information technology systems could generate these feedback reports directly from electronic medical records, then the incremental cost-effectiveness ratio of the provider feedback promotional effort would be $196 per additional person screened for colorectal cancer, which is an amount that is below the midrange of the cost-effectiveness estimates for breast and cervical cancer screening promotional interventions reported previously (Table 1). Investigators at the VA Colorectal Cancer Quality Enhancement Research Initiative have recently developed software programs that extract colorectal cancer screening information directly from electronic medical records (D. Provenzale, personal communication, April 2005).
Our quarterly feedback reports were not entirely automated because we included brief, follow-up training review sessions to maintain dialogue with physicians and to engage in troubleshooting when necessary. This is a core principle of continuous quality improvement, which allows for ongoing evaluation and modification of the intensity and direction of the intervention.23 Prior studies have found that interventions that have relied solely on one-time, automated feedback reports are not effective at improving outcomes over time.30,31 Although our face-to-face review sessions add to the costs of the intervention, this component could be minimized with expanded use of existing technologies to support an interactive, distance education and training format for review sessions.32
The limitations of our study should be identified. First, the findings from our study are based on findings from an urban VA medical center. Additional studies at other VA medical centers, community-based VA outpatient medical centers, and non-VA settings are needed to establish the intervention's generalizability. Because the Medicare program has indicated that it will provide the VA electronic medical record program and software support as a free benefit to all physicians and hospitals in the country, more medical practices will be eligible to incorporate and evaluate our intervention.28 Moreover, improvement initiatives by quality improvement organizations affiliated with the Medicare program are focusing on colorectal cancer screening rates. Our findings suggest that physicians who adopt the VA electronic medical record software could also use related information technology to generate periodic reports of provider-specific colorectal cancer screening rates and potentially improve on them. Second, the use of different cost methodologies across economic studies affects cost-effectiveness findings. For example, Table 1 shows that time spent identifying the target population, overhead, and other indirect costs are included in some but not all cost-effectiveness analyses, making comparability difficult. Third, our study was based on an analysis from the payer perspective, and costs incurred by individual patients were not included in the analysis. For the FOBT, costs to the patient are negligible; the patient can perform the test quickly in his or her home with minimal discomfort. However, flexible sigmoidoscopy and colonoscopy are invasive outpatient procedures that require time spent at the medical center for the procedure, some discomfort, and typically sedation for patients receiving a colonoscopy. Thus, the opportunity costs for these two procedures are much higher. Fourth, longer follow-up of changes in provider behavior is important. Previous intervention studies focused on provider behavior show recidivism over time.33 Finally, it should be noted that a large proportion of patients in the intervention study remained unscreened, and additional efforts may be needed to achieve target goal screening rates.
In conclusion, an intervention that included quarterly feedback reports resulted in a 9% improvement in colorectal cancer screening rates at a VA medical center. The costs associated with this cancer screening promotional effort seem to be reasonable if relevant data can be generated by existing information technology systems. Our findings may have broad applicability because a 2005 Medicare initiative will provide the VA electronic medical record system as a free benefit to all physicians in the country and because progress continues in the refinement of information technology systems in the medical environment.28,34
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
NOTES
Supported in part by Grants No. 1R01CA 102713-01 and P 30 CA60553 from the National Cancer Institute (C.L.B. and J.M.M.) and Grant No. PCI 99-158 from the Department of Veterans Affairs.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
US Preventive Services Task Force: Screening for colorectal cancer: Recommendations and rationale. Ann Intern Med 137:129-131, 2002
Calle EE, Flanders WD, Thun MJ, et al: Demographic predictors of mammography and Pap smear screening in US women. Am J Public Health 83:53-60, 1993
National Cancer Institute Breast Cancer Screening Consortium: Screening mammography: A missed clinical opportunity? Results of the NCI Breast Cancer Screening Consortium and National Health Interview Survey studies. JAMA 264:54-58, 1990
Andersen MR, Urban N: The use of mammography by survivors of breast cancer. Am J Public Health 88:1713-1714, 1998
Andersen MR, Smith R, Meischke H, et al: Breast cancer worry and mammography use by women with and without a family history in a population-based sample. Cancer Epidemiol Biomarkers Prev 12:314-320, 2003
American Cancer Society: Colorectal Cancer Facts and Figures. Atlanta, GA, American Cancer Society, 2005
Centers for Disease Control and Prevention: Colorectal cancer test use among persons aged 50 years–United States, 2001. MMWR Morb Mortal Wkly Rep 52:193-196, 2003
Department of Veterans Affairs: VA Health Care, Systemwide WORKLOAD, FY 1997–2002. http://www.va.gov/vetdata/ProgramStatics/stat_app02/Table%206%20(02).xls
Jha AK, Perllin JB, Kizer KW, et al: Effect of the transformation of the Veterans Affairs Health Care System on the quality of care. N Engl J Med 348:2218-2227, 2003
Ferreira MR, Dolan NC, Fitzgibbon ML, et al: A health-care provider-directed intervention of increase colorectal cancer screening among veterans: Results of a randomized controlled trial. J Clin Oncol 23:1548-1554, 2005
Andersen MR, Urban N, Ramsey S, et al: Examining the cost-effectiveness of cancer screening promotion. Cancer 101:1229-1238, 2004 (suppl 5)
Andersen MR, Hager M, Su C, et al: Analysis of the cost-effectiveness of mammography promotion by volunteers in rural communities. Health Educ Behav 29:755-770, 2002
Stockdale SE, Keeler E, Duan N, et al: Costs and cost-effectiveness of a church-based intervention to promote mammography screening. Health Serv Res 35:1037-1057, 2000
Duan N, Fox SA, Derose KP, et al: Maintaining mammography adherence through telephone counseling in a church-based trial. Am J Public Health 90:1468-1471, 2000
Lynch FL, Whitlock EP, Valanis BG, et al: Cost-effectiveness of a tailored intervention to increase screening in HMO women overdue for Pap test and mammography services. Prev Med 38:403-411, 2004
Thompson B, Thompson LA, Andersen MR, et al: Costs and cost-effectiveness of a clinical intervention to increase mammography utilization in an inner city public health hospital. Prev Med 35:87-96, 2002
Weber BE, Reilly BM: Enhancing mammography use in the inner city, a randomized trial of intensive case management. Arch Intern Med 157:2345-2349, 1997
Fishman P, Taplin S, Meyer D, et al: Cost-effectiveness of strategies to enhance mammography use. Eff Clin Pract 3:213-220, 2000
Taplin SH, Barlow WE, Ludman E, et al: Testing reminder and motivational telephone calls to increase screening mammography: A randomized study. J Natl Cancer Inst 92:233-242, 2000
Mohler PJ: Enhancing compliance with screening mammography recommendations: A clinical trial in a primary care office. Fam Med 27:117-121, 1995
Lantz PM, Stencil D, Lippert MT, et al: Implementation issues and costs associated with a proven strategy for increasing breast and cervical cancer screening among low income women. J Public Health Manag Pract 2:54-59, 1996
Saywell RM, Champion VL, Skinner CS, et al: Cost-effectiveness comparison of five interventions to increase mammography screening. Prev Med 29:374-382, 1999
Shortell SM, Bennett CL, Byck GR: Assessing the impact of continuous quality improvement on clinical practice: What it will take to accelerate progress. Milbank Q 76:593-624, 1998
Russell LB, Gold MR, Siegel JE, et al: The role of cost-effectiveness analysis in health and medicine. JAMA 276:1172-1177, 1996
Weinstein MC, Siegel JE, Gold MR, et al: Recommendations of the Panel on Cost-Effectiveness in Health and Medicine. JAMA 276:1253-1258, 1996
Siegel JE, Weinstein MC, Russell LB, et al: Recommendations for reporting on cost-effectiveness analyses. JAMA 276:1339-1341, 1996
Fitzner KA, Shortridge KF, McGhee SM, et al: Cost-effectiveness study on influenza prevention in Hong Kong. Health Policy 56:215-234, 2001
Kolata G: In unexpected Medicare benefit, US will offer doctors free electronic medical records system. New York Times, July 21, 2005, p A14
Pignone M, Saha S, Hoerger T, et al: Cost-effectiveness analyses of colorectal cancer screening: A systematic review for the U.S. Preventive Services Task Force. Ann Intern Med 137:96-104, 2002
Beck CA, Hugues R, Tu JV, et al: Administrative data feedback for effective cardiac treatment: AFFECT, a clustered randomized trial. JAMA 294:309-318, 2005
Demakis JG, Beauchamp C, Cull WL, et al: Improving residents' compliance with standards of ambulatory care: Results from the VA Cooperative Study on Computerized Reminders. JAMA 284:1411-1416, 2000
McAleer JJ, O'Loan D, Hollywood DP: Broadcast quality teleconferencing for oncology. Oncologist 6:459-462, 2001
Majumdar SR, Soumerai SB: Why most interventions to improve physician prescribing do not seem to work. CMAJ 169:30-31, 2003
Wears RL, Berg M: Computer technology and clinical work: Still waiting for Godot. JAMA 293:1261-1263, 2005(Michael S. Wolf, Karen A.)
Robert H. Lurie Comprehensive Cancer Center and Divisions of General Internal Medicine, Hematology/Oncology, Geriatric Medicine, and Gastroenterology, Department of Medicine, Northwestern University Feinberg School of Medicine
Disease Management Association of America
Veterans Affairs Mid-West Center for Health Services and Policy Research, Jesse Brown Veterans Affairs Medical Center/Lakeside Community-Based Outpatient Clinic
Center for Pharmacoeconomic Research, College of Pharmacy, University of Illinois at Chicago, Chicago, IL.
ABSTRACT
PURPOSE: Colorectal cancer screening is underused, particularly in the Veterans Affairs (VA) population. In a randomized controlled trial, a health care provider–directed intervention that offered quarterly feedback to physicians on their patients' colorectal cancer screening rates led to a 9% increase in colorectal cancer screening rates among veterans. The objective of this secondary analysis was to assess the cost effectiveness of the colorectal cancer screening promotion intervention.
METHODS: Providers in the intervention arm attended an educational workshop on colorectal cancer screening and received confidential feedback on individual and group-specific colorectal cancer screening rates. The primary end point was completion of colorectal cancer screening tests. Sensitivity analyses investigated cost-effectiveness estimates varying the data collection methods, costs of labor and technology, and the effectiveness of the intervention.
RESULTS: Rates of colorectal cancer screening for the intervention versus control arms were 41.3% v 32.4%, respectively (P < .05). The incremental cost-effectiveness ratio was $978 per additional veteran screened based on feedback reports generated from manual review of records. However, if feedback reports could be generated from information technology systems, sensitivity analyses indicate that the cost-effectiveness estimate would decrease to $196 per additional veteran screened.
CONCLUSION: An intervention based on quarterly feedback reports to physicians improved colorectal cancer screening rates at a VA medical center. This intervention would be cost effective if relevant data could be generated by existing information technology systems. Our findings may have broad applicability because a 2005 Medicare initiative will provide the VA electronic medical record system as a free benefit to all US physicians.
INTRODUCTION
The importance of screening for colorectal cancer as a component of evidence-based medicine is well documented.1 Colorectal cancer screening is recommended for average-risk individuals who are 50 years old and older; screening methods include an annual fecal occult blood test (FOBT), a flexible sigmoidoscopy or double-contrast barium enema every 5 years, and a colonoscopy every 10 years.1 However, national screening rates for colorectal cancer are less than 50% and much lower than rates reported for prostate, cervical, and breast cancers.2-5 Individuals who are racial or ethnic minorities and/or who are of lower socioeconomic status have especially low rates of colorectal cancer screening, which may contribute in part to the disproportionately large impact of colorectal cancer on those population subgroups.6,7 This issue is especially relevant in the Veterans Affairs (VA) health care system, the largest integrated delivery system in the country, which provides care for 2% of the US population.8 In 2003, colorectal cancer screening rates were the lowest of the 17 routinely reported performance measures included in the External Peer Review Program of VA medical centers.9
Recently, we reported that a health care provider–directed intervention resulted in a 9% increase in colorectal cancer screening rates at a large urban VA medical center.10 For this intervention to be adopted by others, information is needed on the effectiveness, costs, and cost effectiveness of the intervention. Previous studies have not evaluated the economics of screening promotion efforts for persons who are at average risk for the development of colorectal cancer.11 Studies of individuals at average risk for breast and cervical cancers indicate considerable variability in the cost of health promotion interventions, ranging from less than $10 to more than $2,000 per additional woman screened for each of these two cancers (Table 1). The objective of this study is to evaluate the costs and the cost effectiveness of our health care provider–directed intervention for colorectal cancer among veterans.
METHODS
Provider-Directed Continuous Quality Improvement Intervention to Promote Colorectal Cancer Screening Among Veterans
The Chicago VA Health Care System General Medicine Program operates as three independent firms. In the recent intervention, one firm served as the intervention arm, and all of the physicians in this firm attended an introductory session outlining the importance of colorectal cancer screening as well as tips for efficient communication to veterans, many of whom have poor literacy skills. The intervention was based on a continuous quality improvement framework.23
Approximately every quarter, the providers were invited to attend feedback sessions, during which they received information on the firm's colorectal cancer screening recommendation rate and patient adherence to recommended tests. The providers also received confidential information on their individual recommendation and adherence rates. These sessions included reviews of colorectal cancer screening guidelines and practical strategies to communicate with patients with low literacy skills in busy, high-volume VA primary care clinics. The providers also received quarterly feedback reports on their own colorectal cancer screening recommendation and performance rates and comparative information on the firm-specific rates of these measures.
In the control firm, physicians did not attend the introductory session or receive quarterly feedback reports. Physicians in both firms received the current standard of an electronic, patient-specific reminder through the electronic medical record for colorectal cancer screening eligibility. A research assistant abstracted information on colorectal cancer screening recommendation and performance from the electronic medical records of veterans in the general medicine clinic who were 50 years of age or older, were of average risk for colorectal cancer, and had not returned a three-card FOBT within the previous year or undergone a screening flexible sigmoidoscopy/colonoscopy procedure within the previous 5 years. This information was entered into a relational database. The study statistician generated provider-specific and global firm-specific reports that contained colorectal cancer screening recommendation and performance rates for feedback to health care providers in the intervention arm. Sixty and 53 health care providers cared for 1,015 and 963 eligible study patients for the intervention and controlfirms, respectively.10
Flow Diagram
The flow diagram (Fig 1) represents the sequence of events from the initial random assignment of the firms. This analysis allowed us to identify the relevant clinical outcomes in the intervention and control groups (colorectal cancer screening recommendation and performance) and the resources used in generating the quarterly feedback reports used in the intervention arm.
Assumptions for the Economic Model
We made the following two simplifying assumptions: (1) health care provider participation in the educational/feedback sessions occurred at no cost to the provider because the sessions were conducted over lunchtime and (2) provider-patient discussions about colorectal cancer screening were conducted at no cost because it was opportunistic and required minimal time investment by both parties. These assumptions are consistent with other published cost-effectiveness studies.15-22,24-27
Costs of Obtaining Provider-Specific Colorectal Cancer Screening Rates via Manual Review of Electronic Medical Records
We identified and determined all inputs used for the intervention based on generation of feedback reports by manual review of electronic medical records. The research team (three researchers and two statisticians) provided time estimates for the following: manual review of electronic medical records to identify eligible patients and to assess whether screening had been recommended and completed; data input and analysis for report generation; and development and implementation of the promotional program itself. Costs were derived for the entire 25-month study period. The specific resources included personnel, overhead (amount assigned by institution for normal operating expenses; eg, electricity, which is assumed to be 26% based on VA cost-accounting estimates), and costs related to conducting the feedback sessions. Personnel cost estimates were derived from actual salary and benefit data. A base salary of approximately $30,000 per year for a research assistant was assumed, with a fringe benefit rate of 23.5%. Health care resources not related to implementing the intervention (eg, screening) were excluded (Table 2).
Costs of Generating Provider-Specific Colorectal Cancer Screening Rates via Information Technology Systems
In a second analysis, the data included in the feedback reports for the intervention arm were assumed to be generated with VA information technology support rather than by manual review of electronic medical records. Given the ease of electronic data acquisition within the VA healthcare system, we viewed this to be a logical next step for intervention implementation. Furthermore, a new government program was recently announced that will improve the access of physician practices to electronic medical record technology, increasing the likelihood that feedback could be automated at a larger proportion of non-VA sites in the near future.28 Expert opinion from an information technology specialist, two health economists, and two health services researchers resulted in an estimate of a 90% reduction in personnel time used in the manual review of patients' electronic medical records.
Data Analysis
The cost effectiveness of this provider-directed intervention was evaluated. The incremental cost-effectiveness ratio was derived based on the difference in the costs assigned to the intervention and control groups divided by the difference in colorectal cancer screening rates between the intervention and control groups. Estimates were derived for the incremental cost per additional veteran screened for colorectal cancer (cost effectiveness) and for the average cost of the colorectal cancer screening promotion program per provider involved (mean costs of the intervention per provider). Sensitivity analyses investigated cost-effectiveness estimates varying the source of patient data (derived with information technology review of the electronic medical record v abstracted from the medical record by a research assistant), labor costs, and the effectiveness of the intervention. Because health care policy and funding decisions may involve cost-effectiveness data, the economic analysis was conducted from the third party–payer perspective.12-14 For contextual purposes, the incremental cost-effectiveness ratios (measured as dollars per additional veteran screened) for colorectal cancer screening promotion derived in this intervention were compared with ratios previously reported for other cancer screening promotional efforts (Table 3).
RESULTS
Colorectal cancer screening was recommended for 76.0% of patients in the intervention firm and for 69.4% of controls (P = .02). Rates of colorectal screening (effectiveness data) for the intervention versus control arms were 41.3% v 32.4%, respectively (P = .003).10
Obtaining Provider-Specific Colorectal Cancer Screening Rates via Manual Review of Electronic Medical Records
The total annual cost of the health care provider–directed intervention, using manual review of electronic medical records to generate provider-specific colorectal cancer screening rate data, was $46,741 (Table 2). Over the 25-month intervention period, the total cost of the colorectal cancer screening promotional effort was $86,753, and the mean cost per individual provider was $1,446. The incremental cost-effectiveness ratio for the intervention using manual review of electronic medical records was $978 per additional veteran screened for colorectal cancer.
Sensitivity estimates based on 95% CIs around the effectiveness estimate (38.2% to 52.6%) indicated that the estimated cost per additional veteran screened in the intervention arm ranged from $767 to $3,213. Additional sensitivity analysis estimates based on a 10% variation around the point estimate for costs of manual review of medical records ranged from $78,078 to $95,428 for the 25-month period, and associated incremental cost-effectiveness estimates ranged from $877 to $1,072 per additional veteran screened.
Generating Provider-Specific Colorectal Cancer Screening Rates via Information Technology Systems
We also evaluated the costs of the promotional program assuming that the health care provider–specific colorectal cancer screening rates could be generated directly from information technology supported databases. In this scenario, data collection costs for generating provider-specific feedback rates were estimated to decrease by 90% to $7,702. The overall costs of using information technology–generated feedback data would be $17,431, and the mean costs per provider would be $291. The incremental cost-effectiveness ratio was estimated as $196 per additional person screened for colorectal cancer with feedback data generated by information technology systems.
Corresponding sensitivity estimates for the information technology–supported colorectal cancer screening intervention based on 95% CIs around the point estimate for the effectiveness of the intervention ranged from $156 to $655 per additional veteran screened. Additional sensitivity analysis estimates for the information technology–supported intervention based on a 10% variation around the point estimate for the costs of electronically generating colorectal cancer screening rates ranged from $15,688 to $19,174 for the 25-month period, corresponding to incremental cost-effectiveness estimates of $176 to $215 per additional veteran screened.
DISCUSSION
Our economic analysis provides the first estimates of the costs and cost effectiveness of colorectal cancer screening promotion among veterans at average risk of developing colorectal cancer who receive care at an urban VA medical center. An innovative health care provider–directed intervention, consisting primarily of quarterly feedback sessions highlighting colorectal cancer screening rates of individual providers and average rates of all providers in a particular clinic, resulted in a 9% absolute increase in screening rates over a 2-year period.10 The cost of the feedback program, when manual review of electronic medical records was used to obtain screening rates, was $978 per additional veteran screened for colorectal cancer. However, if the screening data could be generated directly from information technology–supported data systems, the cost could be dramatically reduced to $196 per additional veteran screened.
Our findings can be interpreted in the context of key elements for evaluating the cost effectiveness of cancer screening promotional studies suggested by Andersen et al11 and outlined in Table 3. These concepts address the following seven factors: (1) basing interventions around screening tests that are known to be cost effective; (2) targeting average-risk individuals who have not previously participated in screening efforts; (3) evaluating quality-of-life concerns of participants involved in cancer screening programs; (4) focusing on interventions that are likely to be cost effective; (5) formally assessing the cost effectiveness of the promotion efforts during a randomized intervention; (6) recognizing that the unique characteristics of the patient population, physicians, and health care system play an important role in the promotional efforts; and (7) addressing the sustainability of the promotional program. This health care provider–directed intervention addresses six of the seven proposed lessons, with the exception of including prospective quality-of-life assessments and cost-effectiveness data in terms of quality-adjusted life years. However, the purpose of this analysis was not to elucidate the cost effectiveness of the screening technology (which has been repeatedly investigated)29 but rather to study the promotion strategy itself and to provide further insight into how to improve the cost effectiveness of future colorectal cancer screening promotion strategies.11
Our economic assessment highlights the importance of considering data collection costs when evaluating the feasibility of providing quarterly colorectal cancer screening rate information to health care providers. If information technology systems could generate these feedback reports directly from electronic medical records, then the incremental cost-effectiveness ratio of the provider feedback promotional effort would be $196 per additional person screened for colorectal cancer, which is an amount that is below the midrange of the cost-effectiveness estimates for breast and cervical cancer screening promotional interventions reported previously (Table 1). Investigators at the VA Colorectal Cancer Quality Enhancement Research Initiative have recently developed software programs that extract colorectal cancer screening information directly from electronic medical records (D. Provenzale, personal communication, April 2005).
Our quarterly feedback reports were not entirely automated because we included brief, follow-up training review sessions to maintain dialogue with physicians and to engage in troubleshooting when necessary. This is a core principle of continuous quality improvement, which allows for ongoing evaluation and modification of the intensity and direction of the intervention.23 Prior studies have found that interventions that have relied solely on one-time, automated feedback reports are not effective at improving outcomes over time.30,31 Although our face-to-face review sessions add to the costs of the intervention, this component could be minimized with expanded use of existing technologies to support an interactive, distance education and training format for review sessions.32
The limitations of our study should be identified. First, the findings from our study are based on findings from an urban VA medical center. Additional studies at other VA medical centers, community-based VA outpatient medical centers, and non-VA settings are needed to establish the intervention's generalizability. Because the Medicare program has indicated that it will provide the VA electronic medical record program and software support as a free benefit to all physicians and hospitals in the country, more medical practices will be eligible to incorporate and evaluate our intervention.28 Moreover, improvement initiatives by quality improvement organizations affiliated with the Medicare program are focusing on colorectal cancer screening rates. Our findings suggest that physicians who adopt the VA electronic medical record software could also use related information technology to generate periodic reports of provider-specific colorectal cancer screening rates and potentially improve on them. Second, the use of different cost methodologies across economic studies affects cost-effectiveness findings. For example, Table 1 shows that time spent identifying the target population, overhead, and other indirect costs are included in some but not all cost-effectiveness analyses, making comparability difficult. Third, our study was based on an analysis from the payer perspective, and costs incurred by individual patients were not included in the analysis. For the FOBT, costs to the patient are negligible; the patient can perform the test quickly in his or her home with minimal discomfort. However, flexible sigmoidoscopy and colonoscopy are invasive outpatient procedures that require time spent at the medical center for the procedure, some discomfort, and typically sedation for patients receiving a colonoscopy. Thus, the opportunity costs for these two procedures are much higher. Fourth, longer follow-up of changes in provider behavior is important. Previous intervention studies focused on provider behavior show recidivism over time.33 Finally, it should be noted that a large proportion of patients in the intervention study remained unscreened, and additional efforts may be needed to achieve target goal screening rates.
In conclusion, an intervention that included quarterly feedback reports resulted in a 9% improvement in colorectal cancer screening rates at a VA medical center. The costs associated with this cancer screening promotional effort seem to be reasonable if relevant data can be generated by existing information technology systems. Our findings may have broad applicability because a 2005 Medicare initiative will provide the VA electronic medical record system as a free benefit to all physicians in the country and because progress continues in the refinement of information technology systems in the medical environment.28,34
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
NOTES
Supported in part by Grants No. 1R01CA 102713-01 and P 30 CA60553 from the National Cancer Institute (C.L.B. and J.M.M.) and Grant No. PCI 99-158 from the Department of Veterans Affairs.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
US Preventive Services Task Force: Screening for colorectal cancer: Recommendations and rationale. Ann Intern Med 137:129-131, 2002
Calle EE, Flanders WD, Thun MJ, et al: Demographic predictors of mammography and Pap smear screening in US women. Am J Public Health 83:53-60, 1993
National Cancer Institute Breast Cancer Screening Consortium: Screening mammography: A missed clinical opportunity? Results of the NCI Breast Cancer Screening Consortium and National Health Interview Survey studies. JAMA 264:54-58, 1990
Andersen MR, Urban N: The use of mammography by survivors of breast cancer. Am J Public Health 88:1713-1714, 1998
Andersen MR, Smith R, Meischke H, et al: Breast cancer worry and mammography use by women with and without a family history in a population-based sample. Cancer Epidemiol Biomarkers Prev 12:314-320, 2003
American Cancer Society: Colorectal Cancer Facts and Figures. Atlanta, GA, American Cancer Society, 2005
Centers for Disease Control and Prevention: Colorectal cancer test use among persons aged 50 years–United States, 2001. MMWR Morb Mortal Wkly Rep 52:193-196, 2003
Department of Veterans Affairs: VA Health Care, Systemwide WORKLOAD, FY 1997–2002. http://www.va.gov/vetdata/ProgramStatics/stat_app02/Table%206%20(02).xls
Jha AK, Perllin JB, Kizer KW, et al: Effect of the transformation of the Veterans Affairs Health Care System on the quality of care. N Engl J Med 348:2218-2227, 2003
Ferreira MR, Dolan NC, Fitzgibbon ML, et al: A health-care provider-directed intervention of increase colorectal cancer screening among veterans: Results of a randomized controlled trial. J Clin Oncol 23:1548-1554, 2005
Andersen MR, Urban N, Ramsey S, et al: Examining the cost-effectiveness of cancer screening promotion. Cancer 101:1229-1238, 2004 (suppl 5)
Andersen MR, Hager M, Su C, et al: Analysis of the cost-effectiveness of mammography promotion by volunteers in rural communities. Health Educ Behav 29:755-770, 2002
Stockdale SE, Keeler E, Duan N, et al: Costs and cost-effectiveness of a church-based intervention to promote mammography screening. Health Serv Res 35:1037-1057, 2000
Duan N, Fox SA, Derose KP, et al: Maintaining mammography adherence through telephone counseling in a church-based trial. Am J Public Health 90:1468-1471, 2000
Lynch FL, Whitlock EP, Valanis BG, et al: Cost-effectiveness of a tailored intervention to increase screening in HMO women overdue for Pap test and mammography services. Prev Med 38:403-411, 2004
Thompson B, Thompson LA, Andersen MR, et al: Costs and cost-effectiveness of a clinical intervention to increase mammography utilization in an inner city public health hospital. Prev Med 35:87-96, 2002
Weber BE, Reilly BM: Enhancing mammography use in the inner city, a randomized trial of intensive case management. Arch Intern Med 157:2345-2349, 1997
Fishman P, Taplin S, Meyer D, et al: Cost-effectiveness of strategies to enhance mammography use. Eff Clin Pract 3:213-220, 2000
Taplin SH, Barlow WE, Ludman E, et al: Testing reminder and motivational telephone calls to increase screening mammography: A randomized study. J Natl Cancer Inst 92:233-242, 2000
Mohler PJ: Enhancing compliance with screening mammography recommendations: A clinical trial in a primary care office. Fam Med 27:117-121, 1995
Lantz PM, Stencil D, Lippert MT, et al: Implementation issues and costs associated with a proven strategy for increasing breast and cervical cancer screening among low income women. J Public Health Manag Pract 2:54-59, 1996
Saywell RM, Champion VL, Skinner CS, et al: Cost-effectiveness comparison of five interventions to increase mammography screening. Prev Med 29:374-382, 1999
Shortell SM, Bennett CL, Byck GR: Assessing the impact of continuous quality improvement on clinical practice: What it will take to accelerate progress. Milbank Q 76:593-624, 1998
Russell LB, Gold MR, Siegel JE, et al: The role of cost-effectiveness analysis in health and medicine. JAMA 276:1172-1177, 1996
Weinstein MC, Siegel JE, Gold MR, et al: Recommendations of the Panel on Cost-Effectiveness in Health and Medicine. JAMA 276:1253-1258, 1996
Siegel JE, Weinstein MC, Russell LB, et al: Recommendations for reporting on cost-effectiveness analyses. JAMA 276:1339-1341, 1996
Fitzner KA, Shortridge KF, McGhee SM, et al: Cost-effectiveness study on influenza prevention in Hong Kong. Health Policy 56:215-234, 2001
Kolata G: In unexpected Medicare benefit, US will offer doctors free electronic medical records system. New York Times, July 21, 2005, p A14
Pignone M, Saha S, Hoerger T, et al: Cost-effectiveness analyses of colorectal cancer screening: A systematic review for the U.S. Preventive Services Task Force. Ann Intern Med 137:96-104, 2002
Beck CA, Hugues R, Tu JV, et al: Administrative data feedback for effective cardiac treatment: AFFECT, a clustered randomized trial. JAMA 294:309-318, 2005
Demakis JG, Beauchamp C, Cull WL, et al: Improving residents' compliance with standards of ambulatory care: Results from the VA Cooperative Study on Computerized Reminders. JAMA 284:1411-1416, 2000
McAleer JJ, O'Loan D, Hollywood DP: Broadcast quality teleconferencing for oncology. Oncologist 6:459-462, 2001
Majumdar SR, Soumerai SB: Why most interventions to improve physician prescribing do not seem to work. CMAJ 169:30-31, 2003
Wears RL, Berg M: Computer technology and clinical work: Still waiting for Godot. JAMA 293:1261-1263, 2005(Michael S. Wolf, Karen A.)