A health information technology system designed to facilitate population-based breast cancer screening increased mammography rates in overdue women beyond rates achieved with office-based reminders alone.
To assess the ability of a health information technology system to facilitate population- based breast cancer screening.
Cohort study with 2-year follow-up after a 1-year cluster randomized trial.
Study population was women 42 to 69 years old receiving care within a 12-practice primary care network. The management informatics system (1) identified women overdue for mammograms, (2) connected them to primary care providers using a web-based tool, (3) created automatically generated outreach letters for patients specified by providers, (4) monitored for subsequent mammography scheduling and completion, and (5) provided practice delegates with a list of women remaining unscreened for reminder phone calls. Eligible women overdue for a mammogram during a 1-year study period included those overdue at study start (prevalent cohort) and those who became overdue during follow-up (incident cohort). The main outcome measure was mammography completion rates over 3 years.
Among 32,688 eligible women, 9795 (30%) were overdue for screening (4487 intervention, 5308 control). Intervention patients were somewhat younger, more likely to be non-Hispanic white, and more likely to have health insurance compared with control patients. Adjusted completion rates in the prevalent cohort (n = 6697) were significantly higher among intervention patients after 3 years (51.7% vs 45.8%; P = .002). For patients in the incident cohort (n = 3098), adjusted completion rates after 2 years were 53.8% versus 48.7%, respectively (P = .052).
Population-based informatics systems can enable sustained increases in mammography screening rates beyond rates seen with office-based visit reminders.
(Am J Manag Care. 2012;18(12):821-829)A health information technology system was designed to facilitate population-based breast cancer screening among overdue women in a primary care network.
The US healthcare system is dramatically expanding the use of health information technology as a way to improve the quality and efficiency of care.1,2 In primary care networks, population- based surveillance is being used to identify specific individuals for prevention or disease management interventions. To date most interventions have focused on the use of electronic health records to facilitate care during office-based visits or inpatient hospital admissions.3-5
A novel informatics system to facilitate population-based preventive cancer screening was developed and implemented within a large primary care network.6 Breast cancer screening was chosen because it is the most common cancer in women and the second-most-common cause of cancer-related death,7 because there is scientific evidence supporting screening to decrease breast cancer mortality,8,9 and because many women are not being regularly screened despite broad consensus about the value of screening, especially for postmenopausal women.10,11 The study’s goal was to increase breast cancer screening rates by identifying eligible women overdue for a mammogram and allowing primary care providers to use an informatics tool to quickly review overdue patients and initiate outreach for those selected for contact. The system then automatically mailed reminder letters to the selected patients, tracked mammogram ordering and completion, and facilitated the scheduling of reminder phone calls by practice delegates for women remaining unscreened.
Previous results demonstrate that among women overdue for screening at the start of the study period (prevalent cohort), this system increased breast cancer screening rates over 1 year of follow-up.12 Outcomes for women who became overdue during the 1-year intervention period (incident cohort), representing those just becoming overdue after prior testing or newly eligible for screening based on age criteria, have not been previously reported. Because this incident cohort represents the ongoing population for reminder systems, the current report compares results in incident and prevalent cohorts and assesses the durability of the 1-time intervention benefit over a 3-year period.
METHODSStudy Design and Randomization
The informatics system used in this study, the controlled, cluster randomized trial method, and the primary outcome results over 1 year among individuals who were overdue for screening at the study start are described elsewhere.6,12 A total of 12 primary care practices were allocated to intervention (n = 6) or usual care (n = 6) control groups after stratifying by practice type, the number of eligible patients, baseline mammography rates, and unaffiliated outside facility screening rates. Providers could not be blinded to group assignment. The study was approved by the institutional review board at Massachusetts General Hospital (MGH).
Setting and Participants
The study population consisted of 163,028 individuals seen in the Massachusetts General Primary Care Practice- Based Research Network during the 3 years ending December 31, 2006. All patients were linked to either a specific primary care physician (PCP) or (for patients who could not be linked to a specific physician) to the primary care practice where they received most of their care, using a previously validated algorithm.13,14 This linking ensured that the review of women overdue for breast cancer screening was by the PCPs or practices most directly responsible for each patient’s care.
Eligible study subjects were women 42 to 69 years of age who had no record of mammography in the prior 2 years. This group included women who were overdue as of the intervention start date (March 20, 2007; prevalent cohort) or became overdue during the first year of follow-up (March 20, 2007-March 19, 2008; incident cohort). Patients were excluded if their listed PCP was outside of the MGH network, they had previously undergone bilateral mastectomy, or they had died. All practices used electronic health records that provided visit-based cancer screening reminders.
The informatics tool was implemented in the 6 intervention practices on March 20, 2007, and remained available to providers through March 19, 2010. During the intervention year (through March 19, 2008), providers received reminders to use the tool. After this 1-year period, providers could still use the tool, but they received no additional reminders and the original patient registry was not updated. For intervention providers, the informatics tool consisted of a web page listing their eligible patients linked to the network’s electronic health record.
Physician and Population Manager Role. Separate list views were visible for PCPs for their own patients and for practice-designated population managers (nurses, medical assistants, or nonclinical staff) for patients in each practice not linked to a specific PCP. Physicians and population managers received 3 e-mail reminders (start date, 3 months, 8 months) with a direct link to the population screening web page during the intervention year. A mailed reminder with step-by-step instructions was sent to physicians not yet using the system after 2 months. The web page could also be accessed directly from the hospital’s intranet and included (1) a list of overdue patients, (2) clinically relevant decision support information to help determine whether or not to initiate patient contact, (3) an actionable component to initiate or defer the mammography screening process. If a provider initiated patient contact, a centralized process was started with a letter. Providers could also defer screening (eg, if the patient had previously declined screening after a discussion or had screening done elsewhere) and remove a patient from their list for the remainder of the study. Electronically signed patient letters were sent centrally and included information about the value of screening and how to schedule a mammogram.
Practice Delegate Role. Physicians and population managers were linked with a practice-specific delegate (nonclinical staff or medical assistant) who used his or her own version of the informatics tool to facilitate tracking and scheduling of patients needing contact. Practice delegates were responsible for contacting patients who did not schedule screening on their own. When speaking with patients, delegates could schedule a mammogram by directly accessing the hospital’s radiology ordering system using the informatics tool.
Patient characteristics, mammography reports, and dates of completion were obtained from an electronic clinical and billing data repository.15 Physician characteristics were obtained from the hospital registrar.
The primary outcome was mammography completion rates among patients overdue for screening at the start of the study (prevalent cohort) and among women newly overdue for mammography (incident cohort) during the first study year, comparing intervention and control practices.12 The maximum length of follow-up was 3 years for those in the prevadlent cohort and at least 2 years among those in the incident cohort.
A mammogram was considered to have been completed if there was an electronic report for an imaging test at a network-affiliated institution or if a mammogram was listed in billing data for the patient. Secondary outcomes included time to mammography completion among all overdue patients (prevalent and incident cohorts), censored by cancer diagnosis, death, or end of follow-up. New cancer diagnoses using Partners Healthcare. Cancer Registry data were compared among intervention and control groups.
Baseline patient and physician/practice characteristics were compared between intervention and control groups and between prevalent and incident cohorts using 2-sample t tests or x2 tests, as appropriate. For the primary outcome, adjusted mammography completion rates and 95% confidence intervals were calculated for both the prevalent and incident cohorts at 1, 2, and 3 years of follow-up using Cox proportional hazard models with the robust sandwich covariance matrix estimate to account for clustering while adjusting for potential confounders (PROC PHREG in SAS version 9.2; SAS Institute Inc, Cary, North Carolina). In these models, physicians were considered as the unit of cluster for physician-connected patients and the population manager was considered the unit of cluster for practice-connected patients. To control for differences in patient and practice characteristics among intervention and control practices, patient age, race/ethnicity, insurance status, English language proficiency, practice type (health center vs non—health center), and number of months since last practice visit were included in the models as covariates. All adjusted rates were calculated by holding these covariates at the population mean levels. Unadjusted time to screening completion survival distributions were depicted with Kaplan-Meier curves and compared using a log-rank test. Adjusted hazard ratios comparing intervention with control practices for the entire follow-up period were also reported from the Cox proportional hazards models. The percentages of patients with new cancer diagnoses were compared between intervention and control groups using x2 tests.
There were 64 eligible physicians and 6 practice population managers in the 6 intervention practices and 74 eligible physicians in the 6 control practices. Among intervention providers, 65 of 70 (92.9%) used the system. There were no significant differences between intervention and control practice physicians with regard to age (47.4 vs 46.9 years; P = .78), years since medical school graduation (19.9 vs 19.2; P = .67), or sex (48.4% vs 51.4% were women; P = .86). Two practices in each arm were community health centers. Screening rates at baseline were similar in intervention and control practice groups (79.5% vs 79.3%; P = .73). depicts practice randomization and follow-up.
Among 32,688 eligible women, 9795 (30%) were overdue for screening during the 1-year study period, including 4487 patients in intervention practices and 5308 patients in control practices (). Intervention and control patients were equally likely to be connected to a specific physician (58.9% vs 58.8%) and overdue for screening at the start of the study (prevalent cohort, 67.9% vs 68.8%). Intervention patients were slightly younger, and more likely to speak English, to be non-Hispanic white, to have health insurance, and to have their last clinic visit further in the past than control patients. Compared with patients who were overdue at the start of the study (prevalent cohort), patients who became overdue during 1-year follow-up (incident cohort) were more likely to be connected to a physician (67.6% vs 54.8%; P <.001), to have commercial health insurance (70.9% vs 61.9%; P <.001), and to have been seen more recently for a practice visit (mean 8.6 [8.6 standard deviation (SD)] months vs 13.8 [8.1 SD] months; P <.001).
Providers took action on 3415 of 4487 (76.1%) intervention patients; 2865 (63.9%) were contacted by letter and 550 (12.3%) were deferred. The most common reasons for deferral included test completion at an outside facility (312 [56.7%]) and patient refusal (89 [16.2%]). Among 3045 intervention patients overdue at baseline (prevalent cohort), action was taken on 2629 (86.3%) patients; 2212 (72.6%) were contacted by letter and 417 (13.7%) were deferred. Among 1442 intervention patients who became overdue during the 1-year follow-up period (incident cohort), action was taken on 786 (54.5%) patients; 653 (45.3%) were contacted by letter and 133 (9.2%) were deferred.
Mammography Screening Rates Over Time Percentage of Overdue Population Screened Over 3-Year Follow-up. The percentage of patients in the intervention and control groups who completed screening over the 3-year follow-up period was analyzed separately for prevalent and incident cohorts. Among patients overdue at baseline (prevalent cohort), adjusted completion rates were significantly higher among patients in the intervention group compared with the control group at 1 year (30.1% vs 26.0%; P = .004), 2 years (41.5% vs 36.2%; P = .002), and 3 years (51.7% vs 45.8%; P = .002) of follow-up (). Among patients becoming overdue during the first year (incident cohort), adjusted completion rates were higher but of borderline statistical significance in the intervention group compared with the control group at 1 year (39.8% vs 35.5%; P = .07) and 2 years (53.8% vs 48.7%; P = .052) of follow-up ().
Time to First Completed Mammogram. Time to first completed mammogram over 3 years of follow-up was stratified by whether patients were overdue for screening at the start of the study (prevalent cohort, ) or became overdue during the first year of follow-up (incident cohort, ). Screening rates were higher in the incident cohort than in the prevalent cohort, and in both cohorts intervention patients completed screening sooner than control patients (log-rank P <.001 for both cohorts). Multivariable Cox regression models controlling for potential confounding factors including age, insurance information, race/ethnicity, language spoken, practice type (community health center or not), and months since last practice visit resulted in adjusted hazard ratios demonstrating a benefit in both prevalent and incident cohorts over the follow-up period (prevalent hazard ratio 1.19, 95% confidence interval [CI] 1.07-1.32, P = .001; incident hazard ratio 1.16, 95% CI 1.05-1.28, P = .004). In addition to intervention status, other significant predictors of time to first completed mammogram in multivariable Cox regression modeling included patient age, insurance status, practice type, and months since last practice visit ().
Identification of New Breast Cancers
Although the intervention group had higher rates of mammography screening, the number of breast cancers diagnosed (n = 82) was similar among the intervention and control groups (8.7 vs 8.1 per 1000 eligible women, respectively; P = .75) as well as among patients in the prevalent and incident populations (data not shown).
This study evaluated a novel informatics system delivering integrated population-based preventive care that was designed to supplement and be independent of visit-based reminders. By following patients treated in a cluster randomized trial in a single primary care network, we demonstrated that such a system could increase mammography rates over a 3-year follow-up period. The intervention was more effective in patients who were overdue at the start of the study than in patients who became overdue over the first year of study follow-up. The web-based informatics tool was used by more than 90% of intervention providers, indicating that it was an easy and feasible way to screen patients without needing a face-to-face encounter.
The benefits of reminder systems for improving preventive cancer screening rates have been demonstrated in both visitbased and population-based settings, but most studies have only examined outcomes over short follow-up periods.16-22 We previously reported increased mammography rates over a 1-year period,12 but it is possible that such interventions speed up time to testing completion without raising overall completion rates. By following patients up to 3 years, we demonstrated the durability of the intervention among patients who were overdue at the start of the study. Although differences between intervention and control group rates narrowed over time, patients from intervention practices remained significantly more likely to be screened after 2 or 3 years.
Another limitation of existing studies of reminder systems is that most focus on one-time screening among patients who have failed to be screened over long periods of time (our prevalent cohort).21 However, once reminder systems are started, most patients becoming overdue will be new to screening based on age or will be newly overdue after prior testing. As expected, screening rates differ among patients in incident and prevalent cohorts (Figures 2A and 2B). For reminder systems such as this one that are designed to continue to identify patients as they become overdue over time, their true benefit needs to be evaluated in this incident cohort. We demonstrated that patients in intervention practices who became overdue after the start of the study (incident cohort) were more likely to complete screening than patients in control practices, but the benefit was smaller than that for the prevalent cohort. Factors that may explain these results include higher screening rates among patients in the incident cohort compared with prevalent cohort regardless of intervention status (Figures 2A and 2B). In addition, fewer patients in the incident cohort than in the prevalent cohort were contacted by letter (45.3% vs 72.6%, respectively). Additional reminders to providers over time to screen newly overdue patients are likely necessary to sustain long-term outcomes in this population.
The introduction of this novel informatics system required a fundamental restructuring of the way providers deliver preventive services. Current fee-for-service payment models generally require face-to-face visits and do not provide reimbursement for population-based, visit-independent care.23 Physicians in intervention practices were not compensated for the time they spent reviewing their lists. Although most providers still used the tool, it is possible that decreased use for patients who became overdue after the start of the study reflected the uncompensated nature of panel management outside of an office visit.
The intervention was designed to supplement the primary care network’s current visit-based, electronic health record reminder systems. Future work should assess whether preventive services are best delivered outside of the limited time available during office-based visits. Removing routine but time-consuming tasks from office visits may change the nature of patient-physician face-to-face contact, potentially improving care continuity and providing physicians with the knowledge they need to optimally use such non—visit-based systems.24,25
Although this study involved a large population followed for an extended time period, several limitations warrant comment. By randomizing at the practice rather than patient level, there were small differences in patient characteristics between intervention and control groups that were adjusted for in multivariable models. However, residual unmeasured confounding may have existed even after adjustment. These results cannot be generalized beyond academic primary care networks with well-developed informatics systems, but may represent what can be achieved within a medical home model of care delivery.26 Only 1 screening cycle was assessed for each patient, and only extended follow-up will determine whether ongoing use sustains higher screening rates. Although rates of screening were higher in the intervention group, we did not demonstrate that these higher rates led to more breast cancers being diagnosed. Future studies examining whether informatics interventions decrease morbidity and mortality will require larger patient populations and considerably longer follow-up periods. Finally, future work should assess the additional cost of the intervention relative to the increase in screening observed.
For health information technology to facilitate transformational change in healthcare, current models for delivering care will need to undergo fundamental restructuring. We have demonstrated that a system for preventive breast cancer screening that (1) allows providers to screen their overdue list without a face-to-face visit and identify who should be contacted, (2) tracks those patients for test scheduling and completion, and (3) has practice delegates follow up with those patients remaining overdue can improve screening rates beyond rates seen with visit-based reminders in primary care practice settings. Over 3 years of follow-up, a novel system for breast cancer screening increased mammography rates among intervention patients compared with control patients. These results support the potential value of using integrated informatics tools to help providers deliver care outside of the usual office-based clinical encounter.
This study is registered with ClinicalTrials.gov (NCT00462891).
Author Affiliations: From the General Medicine Division (SJA, JMA, YC, MJB, RWG), Medical Services, and Laboratory of Computer Science (WTL), Massachusetts General Hospital, Harvard Medical School, Boston, MA.
Funding Source: Grant funding was provided by the National Cancer Institute (NCI 1 R21 CA121908) and by institutional support through the Massachusetts General Hospital Primary Care Operations Improvement program.
Author Disclosures: Drs Atlas and Grant are supported by a grant from the Agency for Healthcare Research and Quality (AHRQ R18 HS018161). The other authors (JMA, YC, WTL, MJB) report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.
Authorship Information: Concept and design (SJA, YC, WTL, MJB, RWG); acquisition of data (SJA, JMA); analysis and interpretation of data (SJA, JMA, YC, MJB, RWG); drafting of the manuscript (SJA, JMA, WTL); critical revision of the manuscript for important intellectual content (SJA, JMA, WTL, MJB, RWG); statistical analysis (JMA, YC); provision of study materials or patients (SJA); obtaining funding (SJA, WTL, MJB); administrative, technical, or logistic support (JMA); and supervision (SJA).
Address correspondence to: Steven J. Atlas, MD, MPH, General Medicine Division, Massachusetts General Hospital, 50 Staniford St, Boston, MA 02114. E-mail: firstname.lastname@example.org. Jain SH, Seidman J, Blumenthal D. How health plans, health systems, and others in the private sector can stimulate ‘meaningful use.’ Health Aff (Millwood). 2010;29(9):1667-1670.
2. Stark P. Congressional intent for the HITECH Act. Am J Manag Care. 2010;16(12 suppl HIT):SP24-SP28.
3. Jones SS, Adams JL, Schneider EC, Ringel JS, McGlynn EA. Electronic health record adoption and quality improvement in US hospitals. Am J Manag Care. 2010;16(12 suppl HIT):SP64-SP71.
4. Ornstein SM, Garr DR, Jenkins RG, Rust PF, Arnon A. Computergenerated physician and patient reminders: tools to improve population adherence to selected preventive services. J Fam Pract. 1991;32(1):82-90.
5. Poon EG, Wright A, Simon SR, et al. Relationship between use of electronic health record features and health care quality: results of a statewide survey. Med Care. 2010;48(3):203-209.
6. Lester WT, Ashburner JM, Grant RW, Chueh HC, Barry MJ, Atlas SJ. Mammography FastTrack: an intervention to facilitate reminders for breast cancer screening across a heterogeneous multi-clinic primary care network. J Am Med Inform Assoc. 2009;16(2):187-195.
7. Jemal A, Siegel R, Xu J, Ward E. Cancer statistics, 2010 [published correction appears in CA Cancer J Clin. 2011;61(2):133-134]. CA Cancer J Clin. 2010;60(5):277-300.
8. Community Preventive Services Task Force. Updated recommendations for client- and provider-oriented interventions to increase breast, cervical, and colorectal cancer screening. Am J Prev Med. 2012; 43(1):92-96.
9. Mandelblatt JS, Cronin KA, Bailey S, et al; Breast Cancer Working Group of the Cancer Intervention and Surveillance Modeling Network. Effects of mammography screening under different screening schedules: model estimates of potential benefits and harms [published correction appears in Ann Intern Med. 2010;152(2):136]. Ann Intern Med. 2009;151(10):738-747.
10. Adams EK, Breen N, Joski PJ. Impact of the National Breast and Cervical Cancer Early Detection Program on mammography and Pap test utilization among white, Hispanic, and African American women: 1996-2000. Cancer. 2007;109(2)(suppl):348-358.
11. Smith-Bindman R, Miglioretti DL, Lurie N, et al. Does utilization of screening mammography explain racial and ethnic differences in breast cancer? Ann Intern Med. 2006;144(8):541-553.
12. Atlas SJ, Grant RW, Lester WT, et al. A cluster-randomized trial of a primary care informatics-based system for breast cancer screening. J Gen Intern Med. 2011;26(2):154-161.
13. Atlas SJ, Chang Y, Lasko TA, Chueh HC, Grant RW, Barry MJ. Is this “my” patient? development and validation of a predictive model to link patients to primary care providers. J Gen Intern Med. 2006;21(9):
14. Atlas SJ, Grant RW, Ferris TG , Chang Y, Barry MJ. Patient-physician connectedness and quality of primary care. Ann Intern Med. 2009; 150(5):325-335.
15. Murphy SN, Chueh HC. A security architecture for query tools used to access large biomedical databases. Proc AMIA Symp. 2002:552-556.
16. Baron RC, Melillo S, Rimer BK, et al; Task Force on Community Preventive Services. Intervention to increase recommendation and delivery of screening for breast, cervical, and colorectal cancers by healthcare providers a systematic review of provider reminders. Am J Prev Med. 2010;38(1):110-117.
17. Baron RC, Rimer BK, Breslow RA, et al; Task Force on Community Preventive Services. Client-directed interventions to increase community demand for breast, cervical, and colorectal cancer screening a systematic review. Am J Prev Med. 2008;35(1)(suppl):S34-S55.
18. Brouwers MC, De Vito C, Bahirathan L, et al. What implementation interventions increase cancer screening rates? a systematic review. Implement Sci. 2011;6:111.
19. Sabatino SA, Habarta N, Baron RC, et al; Task Force on Community Preventive Services. Interventions to increase recommendation and delivery of screening for breast, cervical, and colorectal cancers by healthcare providers systematic reviews of provider assessment and feedback and provider incentives. Am J Prev Med. 2008;35(1) (suppl):S67-S74.
20. Sohl SJ, Moyer A. Tailored interventions to promote mammography
screening: a meta-analytic review. Prev Med. 2007;45(4):252-261.
21. Vernon SW, McQueen A, Tiro JA, del Junco DJ. Interventions to promote repeat breast cancer screening with mammography: a systematic review and meta-analysis. J Natl Cancer Inst. 2010;102(14): 1023-1039.
22. Wagner TH. The effectiveness of mailed patient reminders on mammography screening: a meta-analysis. Am J Prev Med. 1998;14(1): 64-70.
23. Berwick DM. Launching accountable care organizations—the proposed rule for the Medicare Shared Savings Program. N Engl J Med. 2011;364(16):e32.
24. Mauksch LB, Dugdale DC, Dodson S, Epstein R. Relationship, communication, and efficiency in the medical encounter: creating a clinical model from a literature review. Arch Intern Med. 2008;168(13): 1387-1395.
25. Weiner SJ, Barnet B, Cheng TL, Daaleman TP. Processes for effective communication in primary care. Ann Intern Med. 2005;142(8): 709-714.
26. American Academy of Family Physicians (AAFP), American Academy of Pediatrics (AAP), American College of Physicians (ACP), American Osteopathic Association (AOA). Joint Principles of the PCMH. http://www.acponline.org/running_practice/pcmh/understanding/what.htm. Published March 2007. Accessed December 28, 2011.