Introduction: This study compared changes of healthcare quality in a Michigan Medicaid population before and after physician adoption of electronic health records (EHRs) via the Meaningful Use (MU) program for selected Healthcare Effectiveness Data and Information Set (HEDIS) quality of care measures.
Methods: Healthcare measures included well-child visits, cancer screening, and chronic illness quality measures. Utilization data were obtained from Medicaid paid claims and encounter data with providers (N=291) receiving their first MU incentive in 2014 and at least one HEDIS-defined outpatient visit with a Michigan Medicaid enrollee. Paired t-tests with a repeated measures design were utilized to analyze the data.
Results: Improvements in quality of infant well-child visits (mean difference = 10.2) and colorectal cancer screening (mean difference = 8.0 percent) were observed. We found no change or slight decreases for the other selected measures.
Conclusion: These outcomes inform the performance and ability of EHRs to improve quality of healthcare standards particularly as technology continues to evolve under the Centers for Medicare & Medicaid Services (CMS) Interoperability and Patient Access final rule.
Keywords: meaningful use, health information technology, electronic health record, clinical quality metrics
Electronic health records (EHRs) have become standard in over 90 percent of physician practices.1 One goal of the Health Information Technology for Economic and Clinical Health (HITECH) Act of 2009 was to improve the efficiency and quality of patient care through increased use of certified EHRs.2-4 The Meaningful Use (MU) component of the HITECH Act incentivized EHR adoption and use by physicians to accomplish specific tasks (e.g., referrals and screenings) and increase patient and provider access to health information.5,6 Now, EHR use has become a clinical standard, and interoperability aims to improve health information technology (HIT) communications for physician and patient. Initially, the Centers for Medicare & Medicaid Services (CMS) provided monetary incentives and to physicians who adopted EHRs.7-9 In turn, physicians were required to demonstrate MU by reporting clinical quality measures (CQMs) to indicate more efficient and equitable patient care processes, and improved patient outcomes.10
Nationwide, providers participated in either the Medicare EHR Incentive Program managed by CMS or the Medicaid EHR Incentive Program managed by each state.11,12 To qualify for an incentive payment under the Medicaid EHR Incentive Program, a provider had to meet one of the following criteria: 1) have a minimum 30 percent Medicaid patient volume; 2) have a minimum 20 percent Medicaid patient volume as a pediatrician; or 3) practice predominantly in a Federally Qualified Health Center or Rural Health Center having a minimum 30 percent patient volume attributable to medically underserved individuals.13
There are very few studies reporting the associations between EHR adoption and healthcare quality in Medicaid beneficiaries, specifically in primary care outpatient settings. Further, there is limited research on the impact of EHR adoption related to promotion of health quality for chronic disease management, recommended prevention and screening services, and medication management.14 Instead, most studies have examined EHRs’ impact on inpatient settings or productivity and efficiency in data quality and management.15,16 We assert that additional research is justified to examine the impact of MU and the adoption of EHRs on the delivery of health services to outpatient Medicaid populations. It is warranted to ascertain that MU achieved quality improvement goals in a variety of settings and for diverse patient populations as the EHR interoperability is in full operation.
This study evaluates the impact of physicians’ participation in MU on select CQMs for Michigan Medicaid enrollees. It builds upon our initial work in a previously published study17 reporting that providers’ participation in MU may have varying degrees of influence on select clinical quality metrics. Although MU attestation has ended, the effectiveness of provider use of EHRs is still highly relevant. Therefore, using a retrospective design, this study examines physician performance on specific CQMs before (pre) and after (post) MU participation on a cohort of Michigan Medicaid enrollees, a vulnerable patient population. The pre-post MU measures studied here are highly relevant to primary care providers and those working to improve care for vulnerable and underserved populations.18
As part of a larger evaluation effort on the impact of HIE/HIT in Michigan, this project was determined to be non-human subject research by the Michigan State University Institutional Review Board. This retrospective analysis of Michigan Medicaid claims and encounter data examined the impact of physicians’ adoption of EHRs on CQM and beneficiaries’ health services. Physicians who first participated in the program in 2014 were selected as the cohort of interest. Select outcome measures for patients attributed to these physicians were then grouped into either pre-MU (2013) or post-MU (2015). Thus, the selected cohort of patients were enrolled in the Michigan Medicaid program continuously in 2013 through 2015.
The Michigan Office of Health Information Technology provided summary data on all eligible MD and DO physicians who participated in its Medicaid MU program for the time period 2011 through 2016. Data elements for providers included payment year, age, name, gender, geographical location, and ethnicity. Then, the Michigan Medicaid data warehouse was accessed to identify: 1) the patient population and 2) the administrative claims data summarized into quality of care metrics by a rate-generating software (Optum® Symmetry®).
Physician MU enrollees with at least one outpatient visit (as defined by Healthcare Effectiveness Data and Information Set (HEDIS)) were accessed via the Michigan Medicaid data warehouse. The primary logic attributed beneficiaries (patients) to the physician based on a plurality of outpatient visits. If two or more physicians had the same number of outpatient visits for an enrollee, a secondary logic was established to break ties based on which provider submitted the most recent claim.
Providers: The study population of providers was restricted to MDs and DOs in Michigan with receipt of their first Medicaid MU incentive program attestation payment in 2014. Providers such as podiatrists, chiropractors, occupational therapists, midwives, etc. were excluded to establish a consistent cohort of providers who could impact primary care sensitive quality measures.
Medicaid Enrollees: This study included enrollees 1) age 64 and younger; 2) with full Michigan Medicaid healthcare coverage and no other insurance or spend down; 3) attributed to one of the study providers in 2014; and 4) qualified for at least one selected quality measure in 2013 or 2015. This rendered a fluctuating member cohort for the pre- and post-measurement periods.
We assessed 12 objective CQMs across three categories of care, including 1) five measures for preventive care visits (well-child visits for infants, primary care visits for younger children, primary care visit for older children, primary care visit for adolescents, and adults’ access to preventive care); 2) three measures for cancer screening (breast, cervical, and colorectal); and 3) four measures for chronic illness (asthma medication management, spirometry test for COPD, HbA1c test for diabetes, and serum creatinine check for hypertension).
We compared providers’ performance on these measures in 2013 (pre-MU) to their performance in 2015 (post-MU). For physician characteristics, we quantified age, gender, practice type (primary versus specialty), and geographic practice location (urban versus rural). For uniformity across 2013 and 2015 data, we utilized a consistent cohort of patients attributed to study physicians in 2014. For enrollee characteristics, we quantified age, gender, geographic location (urban versus rural), and race.
Significant differences (alpha 0.05) between time periods was determined using a paired t-test. To control for potential confounding effects, a repeated measures design was used to account for the fact that observations (outcomes) are not independent and therefore required modeling of correlation structure.19 This method allowed us to model changes in outcomes between 2013 and 2015 while allowing errors to be correlated and adjusting for all covariates. All statistical analyses were conducted using JMP® Pro, Version 13.1.0. (SAS Institute Inc., Cary, NC.)
Provider and Patient Characteristics
Of the 366 MD and DO providers that received their first MU payment in 2014, 291 had at least one HEDIS-defined outpatient visit with a Michigan Medicaid enrollee in 2014 making them eligible for the study. These providers were predominately male (60.2 percent), located in an urban setting (75.5 percent), with a median age of 49 years, and white (57 percent). The Michigan Medicaid beneficiaries (patients) attributed to the 2014 provider cohort were predominantly female (56.5 percent), live in an urban setting (79.9 percent), with a median age of 16 years, and more likely to be white (54.1 percent). These provider and patient demographic characteristics are displayed in Table 1. Demographic proportions in pre/post years, 2013 and 2015, were similar to those during the attribution year except for patients’ age. Age difference was expected and recognized as an artifact of the study design where those in the pre-study period would be younger. Additionally, age, gender, and enrollment restrictions applied to individual CQMs in accordance with national specifications may be responsible the observed age difference.
Overall comparisons and mean difference (MD) for each measure for the same provider in 2013 (pre) and 2015 (post) revealed significant increases (p<0.05) in three CQMs: 1) well-child visits (six or more) for infants (MD=10.2); 2) primary care visits for adolescents (MD=1.7); and 3) colorectal cancer screening (MD=8.0). Conversely, four measures: 1) primary care visits for younger children (MD= -3.0); 2) primary care visits for older children (MD= -3.0); 3) breast cancer screening (MD= -3.7); and 4) cervical cancer screening (MD= -7.3) demonstrated significant (p<0.05) decline between 2013 and 2015 (Table 2).
Significant differences were noted for some CQMs based on physician characteristics. For female providers, we identified a significant decrease for primary care visits for the adolescent measure [MD = -2.4; 95% CI (-4.1 – -0.7)], and significant increase in the spirometry test for the COPD measure [MD = 12.4; 95% CI (3.5–21.2)]. Providers who practiced in rural areas had a significant increase for well-child visits for infants [MD = 10.3; 95% CI (4.2–16.4)], adults’ access to preventive care [MD=2.2 (0.5–3.8)], and HbA1c test for persons with diabetes [MD=7.4; 95% CI (3.7– 11.2)]. Finally, we identified significant inverse relationships between the providers’ age and two measures: primary care visit for older children [MD = -0.1; 95% CI (-0.2–0.0)] and spirometry test for COPD [MD = -0.9; 95% CI (-1.6 – -0.2)]. The remaining measures showed no significant difference between 2013 and 2015 for measures dependent on provider characteristics. Results are shown in Table 2. For several CQM, we note improvements for rural geographic characteristics attributed to EHR-MU.
Results showed the use of EHR-MU garnered varied outcomes for patients across multiple categories of care. Improvements for infant and adolescent well-child visits were observed, but decreases in childhood well-child visits were shown. These findings could be a function of fewer well-child visits in middle childhood as documented in the literature.20 Thus, EHRs could be used to increase consistent and sustained well-visits across the childhood developmental milestones to improve child health overtime. Given that EHRs have identified this as clinic/system gap, interventions should be implemented to increase child well-visits.
Likewise, cancer screening outcomes showed mixed results with improvements in colorectal cancer screening and decreases in breast and cervical cancer screening attributed to EHR-MU. It should be noted that breast and cervical cancer screening guidelines changed during the study period recommending fewer screenings for low-risk patients. Thus, the decrease in breast and cervical cancer screening rates could be a considered improvement as a result of adherence to the new guidelines but warrants further investigation. The study revealed no changes in chronic disease management, again, indicating a need for provider attention regarding utilization of EHRs to improve patient management. Chronic disease management is complex and requires multifactorial input including patient education on the importance of regular visits, improved provider-patient communication, and utilization of the EHR (physicians as well as patient portals) to increase follow-up rates that are imperative for chronic disease management.
Although we know that this cohort of physicians first participated in the Michigan Medicaid MU Incentive program in 2014, we do not know if they had prior experience or administrative support in using EHRs. This is an important consideration because studies suggest that provider initiation to EHRs can affect efficiency and performance21,22 as well as contributing to physician burnout23-25 that could account for inconsistencies in observed screening and prevention outcomes here. Inconsistent findings have important implications for evaluating the success of the EHR utilization on quality of patient care, particularly for Medicaid populations that often experience health disparities.26,27 Although improvements for infant well-child visits, adult preventive care, and HbA1c test for persons with diabetes were observed for rural geographic characteristics, further examination is needed to make a full interpretation of the outcome.
Limitations of the study include restriction to physicians who participated in the Michigan Medicaid MU Incentive Program and to Michigan Medicaid participants who received care from these physicians, which limits generalization to broader populations who are enrolled in Medicare or private insurers. In addition, CQMs were assessed using only claims data, so there could have been improvements in outcomes not fully captured from these data. Furthermore, we did not have information about practice workflow and provider EHR education that could influence provider MU performance. Finally, we limited our comparison to one year before and after adoption. Additional years of follow-up are necessary to fully identify improvements in care linked to EHR adoption.28 We plan to address this in future studies.
As supported by previous studies in the literature, EHRs play to mixed reviews.29-32 For this Medicaid population in Michigan, improvement in healthcare delivery was found for infant and adolescent well-child visits and colorectal cancer screening but apparent decreases in childhood well-child visits and breast and cervical cancer screening. No significant changes were observed for chronic disease. While our results did not have ability to explain patient outcomes, other sources have suggested that improving EHR software design and updates, as well as effective physician training, would make improvements in patient quality of care more consistent.33 Given that EHRs are now considered a standard of clinical care process, it is necessary to keep physician capacity in mind (e.g., workflow, burnout). Outcomes of this study can shed light on effective physician EHR use and improving patient outcomes.34,35
1. Myrick KL, Ogburn DF, Ward BW. “Percentage of Office-Based Physicians using any Electronic Health Record (EHR)/electronic medical record (EMR) system and Physicians that have a Certified EHR/EMR system by U.S. state: National Electronic Health Records Survey.” National Center for Health Statistics; January 2019. Accessed Dec 2020 at; https://www.cdc.gov/nchs/data/nehrs/2017_NEHRS_Web_Table_EHR_State.pdf
2. Blumenthal D, Tavenner M. “The ‘Meaningful Use” Regulation for Electronic Health Records.” NEJM. 2010;363(6):501-504. doi: 10.1056/NEJMp1006114
3. Burde H. “The HITECH Act: An Overview.” AMA J Ethics. 2011;3(3):172-175. doi: 10.1001/virtualmentor.2011.13.3.hlaw1-1103
4. U.S. Ways and Means Committee. “Health information Technology for Economic and Clinical Health (HITECH) Act.” Code of Federal Regulations; 2009. Accessed Dec 2020 at: https://www.asha.org/Practice/reimbursement/hipaa/HITECH-Act/
5. Buntin MB, Jain SH, Blumenthal, D. “Health Information Technology: Laying the Infrastructure for National Health Reform.” Health Affairs. 2010;29(6):1214-1219. doi: 10.1377/hlthaff.2010.0503
6. Gold M, McLaughlin C. “Assessing HITECH Implementation and Lessons: Five Years Later.” Milbank Q. 2016;94(3):654-687. doi: 10.1111/1468-0009.12214
7. Centers for Medicare & Medicaid Services. Tip Sheet: Medicaid Electronic Health Record Incentive Payments for Eligible Professionals; 2013. Accessed Dec 2020 at: https://www.cms.gov/Regulations-and-Guidance/Legislation/EHRIncentivePrograms/Downloads/MLN_MedicaidEHRProgram_TipSheet_EP.pdf
8. Centers for Medicare & Medicaid Services. “Annual Updates to eCQM Specifications.” Accessed March 2021 at: https://www.cms.gov/Regulations-and-Guidance/Legislation/EHRIncentivePrograms/eCQM_Library.html
9. The Office of the National Coordinator for Health Information Technology. “Medicare and Medicaid EHR Incentive Programs Meaningful Use Core Objectives that Address Privacy and Security.” Privacy and Security of Electronic Health Information. 2015;5:32-34. Access March 2021 at: https://www.healthit.gov/sites/default/files/pdf/privacy/privacy-and-security-guide-chapter-5.pdf
10. Centers for Medicare & Medicaid Services. “Annual Updates to eCQM Specifications.” Accessed March 2021 at: https://www.cms.gov/Regulations-and-Guidance/Legislation/EHRIncentivePrograms/eCQM_Library.html
11. Agency for Healthcare Research and Quality. “Practice Facilitation Handbook Module 17. Electronic Health Records and Meaningful Use.” Agency Health Res Qual, Rockville, MD; 2013. Accessed March 2021 at: https://www.ahrq.gov/ncepcr/tools/pf-handbook/mod17.html
12. Michigan Department of Health & Human Services. “Medicaid EHR Incentive Program: Eligible Professional’s Guide to the Michigan Medicaid EHR Incentive Program” (version 6.1); 2017. Accessed Jun 2021 at: https://michiganhealthit.org/wp-content/uploads/EP-Guide-3.2-2014.pdf
13. Centers for Medicare & Medicaid Services. Tip Sheet: Medicaid Electronic Health Record Incentive Payments for Eligible Professionals; 2013. Accessed Dec 2020 at: https://www.cms.gov/Regulations-and-Guidance/Legislation/EHRIncentivePrograms/Downloads/MLN_MedicaidEHRProgram_TipSheet_EP.pdf
14. Angier H, Jacobs EA, Huguet N, et al. “Progress Towards using a Community Context with Clinical Data in Primary Care.” Fam Med Community Health. 2018;7(1):e000028. doi: 10.1136/fmch-2018-000028.
15. Afonso AM, Alfonso S, Morgan TO. “Short-term Impact of Meaningful Use Stage 1 Implementation: A Comparison of Health Outcomes in Two Primary Care Clinics. J Amb Care and Manag. 2017;40(4), 316-326. doi: 10.1097/JAC.0000000000000179
16. Jung HY, Unruh MA, Vest JR, et al. “Physician Participation in Meaningful Use and Quality of Care for Medicare Fee-for-Service Enrollees.” J Am Geriat Soc. 65(3):608-613. DOI: 10.1111/jgs.14704
17. Brooks K, Sarzynski E, Houdeshell-Putt L, et al. “Meaningful Use: Does Physician Participation Move the Needle on Quality Metrics? J Heal Qual. 2019;41(6):e70-e76. doi: 10.1097/JHQ.0000000000000210
18. Sandefer RH, Marc DT, Kleeberg P. “Meaningful Use Attestations among US Hospitals: The Growing Rural -Urban Divide.” Perspectives in Health Information Management. Spring 2015.
19. Cole JWL, Grizzle JE. (1966). “Applications of Multivariate Analysis of Variance to Repeated Measurement Experiments.” Biometrics. 1966;22:810-828.
20. Wolf ER, Hochheimer CJ, Sabo RT, et al. “Gaps in Well-Child Care Attendance among Primary Care Clinics Serving Low-income Families.” Pediatrics. 2018;142(5):e20174019. DOI: 10.1542/peds.2017-4019
21. Brooks K, Sarzynski E, Houdeshell-Putt L, et al. “Meaningful Use: Does Physician Participation Move the Needle on Quality Metrics? J Heal Qual. 2019;41(6):e70-e76. doi: 10.1097/JHQ.0000000000000210
22. Overhage JM, McCallie D. “Physician Time Spent Using the Electronic Health Record during Outpatient Encounters.” Annals of Internal Medicine, 2020;172:169-174. doi: 10.7326/M18-3684
23. Davis MJ. “Using Technology to Combat Clinician Burnout.” J Health Manag. 2020;65 (4):265-272. DOI: 10.1097/JHM-D-20-00099
24. Heisey-Grove DM, Wall HK, Wright JS. “Electronic Clinical Quality Measure Reporting Challenges: Findings from the Medicare EHR Incentive Program’s Controlling High Blood Pressure Measure.” JAMIA. 2018;25(2):127-134. doi: 10.1093/jamia/ocx049
25. Melnick ER, Dyrbye LN, Sinsky CA, et al. “The Association between Perceived Electronic Health Record Usability and Professional Burnout among US Physicians.” Mayo Clin Proc. 2019; pii:S0025-6196(19)30836-5. DOI: https://doi.org/10.1016/j.mayocp.2019.09.024
26. Bradley CJ, Given CW, Roberts C. “Health Care Disparities and Cervical Cancer.” Am J Public Health, 94:2098-2103.
27. Choi SK, Adams S A, Eberth JM, et al. “Medicaid Coverage Expansion and Implications for Cancer Disparities”. Am J Pub Heal. 2015;105:5706-5712. doi: 10.2105/AJPH.2015.302876
28. Adler-Milstein J. “Electronic Health Record Time Among Outpatient Physicians: Reflections on the Who, What and Why.” Ann of Intern Med. 2020;172:212-213. doi: 10.7326/M19-3921
29. Overhage JM, McCallie D. “Physician Time Spent Using the Electronic Health Record during Outpatient Encounters.” Annals of Internal Medicine, 2020;172:169-174. doi: 10.7326/M18-3684
30. Heisey-Grove DM, Wall HK, Wright JS. “Electronic Clinical Quality Measure Reporting Challenges: Findings from the Medicare EHR Incentive Program’s Controlling High Blood Pressure Measure.” JAMIA. 2018;25(2):127-134. doi: 10.1093/jamia/ocx049
31. American Medical Association. “Improving Care: Priorities to Improve Electronic Health Record Use. Executive Summary”; 2014. Accessed March 2021 at: ama-assn.org/sites/ama-assn.org/files/corp/media-browser/member/about-ama/ehr-priorities.pdf
32. Emani S, Ting DY, Healey M, et al. “Physician Beliefs about the Meaningful Use of the Electronic Health Record: A Follow-Up Study.” Applied clinical Informatics. 20178(4), 1044–1053. https://doi.org/10.4338/ACI-2017-05-RA-0079
33. Colligan L, Sinsky C, Goeders L, et al. “Sources of Physician Satisfaction and Dissatisfaction and Review of Administrative Tasks in Ambulatory Practice: A Qualitative Analysis of Physician and Staff Interviews.” 2020. Accessed March 2021 at: ama-assn.org/go/psps
34. The Office of the National Coordinator for Health Information Technology. “Medicare and Medicaid EHR Incentive Programs Meaningful Use Core Objectives that Address Privacy and Security.” Privacy and Security of Electronic Health Information. 2015;5:32-34. Access March 2021 at: https://www.healthit.gov/sites/default/files/pdf/privacy/privacy-and-security-guide-chapter-5.pdf
35. Centers for Medicare and Medicaid Services. “Policies and Technology for Interoperability and Burden Reduction.” 2021. Accessed May 2021: https://www.cms.gov/Regulations-and-Guidance/Guidance/Interoperability/index#CMS-Interoperability-and-Patient-Access-Final-Rule