Impact of Medical Director Certification on Nursing Home Quality of Care

      Objective

      This study tests the research hypothesis that certified medical directors are able to use their training, education, and knowledge to positively influence quality of care in US nursing homes.

      Design

      F-tag numbers were identified within the State Operations Manual that reflect dimensions of quality thought to be impacted by the medical director. A weighting system was developed based on the “scope and severity” level at which the nursing homes were cited for these specific tag numbers. Then homes led by certified medical directors were compared with homes led by medical directors not known to be certified.

      Data/participants

      Data were obtained from the Centers for Medicare & Medicaid Services' Online Survey Certification and Reporting database for nursing homes. Homes with a certified medical director (547) were identified from the database of the American Medical Directors Association.

      Measurements

      The national survey database was used to compute a “standardized quality score” (zero representing best possible score and 1.0 representing average score) for each home, and the homes with certified medical directors compared with the other homes in the database. Regression analysis was then used to attempt to identify the most important contributors to measured quality score differences between the homes.

      Results

      The standardized quality score of facilities with certified medical directors (n=547) was 0.8958 versus 1.0037 for facilities without certified medical directors (n=15,230) (lower number represents higher quality). When nursing facility characteristics were added to the regression equation, the presence of a certified medical director accounted for up to 15% improvement in quality.

      Conclusions

      The presence of certified medical directors is an independent predictor of quality in US nursing homes.

      Keywords

      Since the introduction of the concept of nursing home medical directors in the 1970s there have been multiple papers, guidelines, and books published on the role of the medical director and how this should affect the quality of care in the nursing home.
      There has, however, been no clear quantification of the impact that a well-trained medical director can have on the quality of care within a facility.
      The official mission statement of the American Medical Directors Certification Program (AMDCP) is to “…advance physician leadership…thereby enhancing quality of care.”

      AMDA. Certified Medical Director in Long Term Care (AMDA CMD). Available at: http://www.amda.com/certification/overview.cfm. Accessed June 17, 2009.

      Since its inception in 1991, the AMDCP has certified more than 2500 medical directors. The certification process follows an “experiential” model that incorporates existing mechanisms such as fellowship programs, board certification, continuing medical education programs (offered by major provider organizations), courses in medical direction (approved by AMDCP), and other continuing education programs. Familiarity with the medical director certification process leads to the expectation that medical director certification is positively correlated with quality of care. Although such a correlation is commonly and reasonably asserted, we have found nothing in the literature empirically demonstrating such a relationship.
      This study tests the hypothesis that certified medical directors are able to use their training, education, and knowledge to positively influence quality of care in US nursing homes. The alternate hypothesis (or null hypothesis) is that certification makes no appreciable difference to nursing home quality of care.

      Methodology

      F tags from the State Operations Manual

      Interpretive Guidelines for Long-Term Care Facilities. Available at: http://cms.hhs.gov/manuals/Downloads/som107ap_pp_guidelines_ltcf.pdf. Accessed June 17, 2009.

      (N=27) were identified that appear to reflect dimensions of quality potentially directly impacted by the medical director. These were chosen by consensus of the research team including the authors, AMDCP staff, and the AMDCP Executive Committee on the premise that these were areas of quality potentially influenced by medical director activity. A weighting scheme based on the “scope and severity” level at which the nursing homes were cited for these specific tag numbers was also developed. This was constructed to emphasize serious, widespread, or patterns of deficiencies. It was thought that a better prepared medical director would be able to reduce the incidence of deficiency citations for these 27 specific F tags, or, at a minimum, reduce the scope and severity level at which they were cited. A listing of the 27 F tags and our scope and severity–based weighting scheme is listed in Appendix 1. The weighting scheme is similar, though not identical, to that used by the Nursing Home Compare Five-Star Rating guide.

      NursingHome Compare. Design for Nursing Home Compare Five-Star Quality Rating System: Technical Users Guide. January 2009. Available at: http://www.cms.hhs.gov/CertificationandComplianc/13_FSQRS.asp. Accessed June 17, 2009.

      Using the Centers for Medicare & Medicaid Services' Online Survey Certification and Reporting (OSCAR) database as of March 2008, a “raw quality score” was computed, and a “standardized quality score” for all 15,777 certified nursing homes that were in operation in the United States in March 2008.

      Obtained from the files of Cowles Research Group. Available at: http://www.longermcareinfo.com/about_oscar.html. Accessed June 17, 2009.

      The raw quality score was computed by summing the weights of the relevant deficiency citations. We then divided the raw quality score by the state average raw quality score to yield a standardized quality score. Standardization of the quality score is necessitated by wide state-to-state variation in the survey process. For example, New Jersey nursing home surveys result in an average of 4 total deficiencies per survey, whereas in neighboring Delaware the comparable average is 13.
      • Cowles C.M.
      Nursing Home Statistical Yearbook: 2007.
      Dividing the raw quality score by the state average “standardizes” the score, creating a measure that is comparable across states. Note that lower adjusted quality scores denote better quality, and an adjusted quality score of unity denotes average quality.
      Records of the American Medical Directors Association (AMDA) were then used to identify 547 nursing homes that had certified medical directors during the year immediately preceding and during the survey contained in our data capture. The first step was to compare the average standardized quality score in facilities with certified medical directors to those without certified medical directors. A “t test” was then computed to evaluate the degree to which the difference between the averages was statistically significant. Finally, other variables were considered that could also affect quality, and multiple regression analysis was used to better understand the relationship between medical director certification and quality of care.
      This project was granted institutional review board (IRB)-exempt status by the IRB of Wright State University via the American Medical Directors Research Foundation. The project was sponsored by generous grants from the American Medical Directors Certification Program and AMDA state chapter contributions.

      Findings

      As shown in Table 1, the average standardized quality score (SQS) in facilities with certified medical directors was 0.8958 compared with 1.0037 for facilities without certified medical directors. Recall that lower numbers represent better quality, with zero representing the best possible score. The difference of 0.1079 represents a 12% improvement in quality associated with the presence of a certified medical director, and the t test indicated that the difference is statistically significant at the 98% level.
      Table 1Average Standardized Quality Score With and Without a Certified Medical Director (CMD)
      With CMD (n=547)0.8958
      Without CMD (n=15,230)1.0037
      Difference #0.1079
      Difference %12.05
      It was suspected, however, that other nursing home characteristics are correlated with quality, such as facility size, class of ownership, case mix, staffing, and urban/rural status. Smaller facilities should tend to have fewer deficiencies because of fewer opportunities for errors, and not-for-profit facilities are known to have better surveys than for-profit facilities.
      • Cowles C.M.
      Nursing Home Statistical Yearbook: 2007.
      The higher case mix associated with more medically complex cases might result in more deficiencies, higher staffing would be expected to result in fewer deficiencies, and rural facilities might have better surveys than urban facilities. The urban/rural impact on quality, if there is one, might more accurately be associated with size or staffing differences between urban and rural nursing homes.
      There appears to be significant potential for the statistical relationships hypothesized in the preceding paragraph to confound the initial findings reported in Table 1. For example, what if facilities with certified medical directors are more likely to be small, or more likely to be not-for-profit? If that were the case, then the variation in quality of care that is attributed to medical director certification in Table 1 might in fact be attributable to these other factors. Stepwise multiple regression analysis was used to help determine if this might be the case and to better understand the relationship between quality and medical director certification. Data from the best specified equation are reported in Table 2. It was found that the strongest predictors of adjusted quality were whether or not the nursing home had a certified medical director, whether or not the total number of beds in the facility was greater than 99, whether or not it was a proprietary (for-profit) facility, and the number of registered nurse (RN) staffing hours per patient day.
      Table 2Regression Equation Predicting Standardized Quality Score
      Dependent Variable: Adjusted Quality Score
      Number of observations read15777
      Number of observations used15618
      Number of observations with missing values159
      Analysis of Variance
      SourceDFSum of SquaresMean SquareF ValuePr>F
      Model4562.03592140.5089858.61<.0001
      Error15613374312.39739
      Corrected total1561737993
      Root MSE1.54835R-Square0.0148
      Dependent mean1.00559Adj R-Square0.0145
      Coefficient of variation153.97489
      Parameter Estimates
      VariableDFParameter EstimateStandard ErrorT ValuePr>/t/
      Intercept10.987800.0346128.54<.0001
      Certified medial director present1−0.147050.06783−2.17.0302
      Beds >9910.223090.025468.76<.0001
      For profit10.089870.027443.7.0011
      Registered nurse hours per patient day1−0.249870.02918−8.56<.0001
      Recall that the average adjusted quality score is 1.0000 and that lower numbers reflect better quality. Thus, the estimated regression coefficient of –0.14705 for the certified medical director variable indicates that, holding other predictors constant, the presence of a certified medical director will improve quality by about 15%. Other results from the regression equation were consistent with stated preconceptions, ie, larger facilities, proprietary facilities, and facilities that staff fewer RN hours per patient day tend to have poorer quality. Urban/rural status and case mix are not included in the equation as they did not improve equation specification.

      Discussion

      For the first time, this study demonstrates that the certified medical director has a measurable positive effect on the quality of care provided in facilities in which they serve.
      In 1975, regulations were promulgated that required skilled nursing facilities to have a medical director. In response to this, AMDA was formed in 1977 to organize the medical directors and provide a venue for education of the medical directors in their role and responsibilities. Numerous articles (representative articles in references) have been written since that time about the role of the medical director.
      • American Medical Directors Association
      Roles and responsibilities of the medical director in the nursing home: position statement A03.
      • Schnelle J.F.
      Total quality management and the medical director.
      • Schnelle J.F.
      • Ouslander J.G.
      CMS guidelines and improving continence care in nursing homes: The role of the medical director.
      • Zimmer J.G.
      • Watson N.M.
      • Levenson S.A.
      Nursing home medical directors: Ideals and realities.
      • Smith R.L.
      • Osterweil D.
      The medical director in hospital-based transitional care units.
      Articles have been written on specific problems in which the medical director can and should make a difference.
      • Colon-Emeric C.S.
      • Casebeer L.
      • Saag K.
      • et al.
      Barriers to providing osteoporosis care in skilled nursing facilities: perceptions of medical directors and directors of nursing.
      • Richards Jr., C.L.
      Preventing antimicrobial-resistant bacterial infections among older adults in long-term care facilities.
      • Munir J.
      • Wright R.J.
      • Carr D.B.
      A quality improvement study on calcium and vitamin D supplementation in long-term care.
      Textbooks on the role of the medical director
      • Pattee J.J.
      • Otteson O.J.
      Medical Direction in the Nursing Home: Principles and Concepts for Physician Administrators.
      and long-term care medicine
      • Ouslander J.G.
      • Osterweil D.
      • Morley J.
      Medical Care in the Nursing Home.
      have been published. To the best of our knowledge, only one has made an attempt to show via survey of medical directors and administrators in Maryland that requiring medical director training makes a positive difference in the quality of medical directorship provided.
      • Boyce B.F.
      • Bob H.
      • Levenson S.A.
      The preliminary impact of Maryland's medical director and attending physician regulations.
      In that study, which reports on a survey of medical directors and administrators following the institution of mandatory medical director education, there was consensus that the relationship between the medical director and other administrators in the nursing home was improved, that the medical director spent more time in the facility working on system issues, and that the medical director spent more time with the administrator reviewing the care provided. In contrast, the present study uses a comparison of actual survey data from the facilities.
      The current study was initiated in an attempt to demonstrate whether the presence of a certified medical director made a measurable difference in the quality of care provided within long-term care facilities. The results support the conclusion that the presence of a certified medical director makes an appreciable and positive difference on the quality of care provided within long-term care facilities. The data also support the premise that there are other important factors determining the quality of care provided.

      Other Factors Affecting Results

      There is great confidence that all 547 nursing homes that were flagged as having certified medical directors during the study period actually did have certified medical directors because all certified medical directors identified their facility as where they worked in their medical director role for AMDA records in the time frame immediately before this study. The comparison group of 15,230 facilities that are treated as not having a certified medical director may actually contain facilities that may have had a certified medical director during all or part of the study period. This is likely, because of the 2500 certifications awarded, it is estimated via AMDA records that approximately 1500 of these individuals are still working. If the truth is that certified medical directors are associated with higher quality, then including facilities in the control group that actually had certified medical directors would tend to reduce the difference between the 2 comparison groups. Thus, if there were inadvertently included facilities in the comparison group that had certified medical directors, then the true difference between the certified and noncertified groups was larger than what is reported, ie, the research results are even more robust than what is reported.
      The probability values reported in Table 2 are for a 2-tailed t test. We could argue on theoretical grounds that the 1-tailed test is more appropriate. Interpreting our t ratios using a 1-tailed test would also make our results more robust, ie, double the level of statistical significance for each predictor variable.
      Of note on the statistical analysis of the linear regression model is that the multiple correlation coefficient (R squared) is relatively low (0.0148); however, it needs to be placed in the context that the goal of this study was to test whether the presence of a certified medical director made a positive impact on the quality of care in that nursing home, not to explain the total variation in the quality measured. Thus, the magnitude of the partial correlation coefficient associated with the certified medical director variable (–0.14705) and its associated level of statistical significance (.0302) are of much greater importance than the absolute value of the multiple correlation coefficient.
      Other factors that theoretically could bias the outcome are that 2 of the authors are currently certified medical directors and medical directors of facilities included in the database, however it is doubtful that 2 individual homes would bias the overall results in comparison with either the 547 identified facilities with a certified medical director or the 15,230 other facilities. A potentially more important variable is that many AMDA members have trained in geriatric fellowship programs and have certification in geriatric medicine. Of the certified medical director–led facilities in this study, 18% (101 of the 547) are led by medical directors with geriatric fellowship training. We did not attempt to separate out the contribution of this training in the current project, but a recent survey study examined barriers to care and visit time expectations, which revealed that geriatric-trained physicians may have a higher level of expectation in their care of long-term care patients.
      • Caprio T.V.
      • Karuza J.
      • Katz P.R.
      Profile of physicians in the nursing home: Time perception and barriers to optimal medical practice.
      Currently, all long-term care facilities are required to have a physician identified as medical director. The data now reported suggest that there is a clear and measurable positive effect on quality if that medical director is a certified medical director. This may have policy implications in all of long-term care. Because the certified medical director designation indicates a minimum level of experience and education in medical director management and clinical geriatric medicine, it suggests that every long-term facility and program should have a certified medical director or the equivalent. An alternate explanation is that certified medical directors are a self-identified group of dedicated, experienced individuals who are willing to be held accountable as long-term care providers and leaders, and that they would be so whether or not they had attained recognition as a certified medical director. Whatever the reason, our patients deserve the best of all of us.

      Conclusion

      This research demonstrates that the presence of a certified medical director in a facility makes an appreciable positive difference in the quality of care provided in that facility. The data also identify other factors—small facility size, not-for-profit status, and higher RN hours per patient day—as important determinants of higher quality offered by a facility. It is hoped that this will lead to further recognition of the knowledge and skills of trained medical directors, and encourage all medical directors to work to attain and improve these skills.

      Appendix 1

       F-tags Included in Standard Quality Scores

      Tabled 1
      F-TagArea of Medical Direction
      202Identify appropriate ways to minimize avoidable transfers
      221–222Restraints: Policies and procedures; alternatives to use
      223Freedom from abuse
      280Train attending physicians to help staff develop resident care plan
      281–282Medical direction—additional duties
      309QA relative to MDS
      314Pressure ulcers
      319–320Access to mental health treatment
      323–324Minimizing and reporting accidents
      325Weight loss and nutrition
      329Unnecessary drugs
      385, 386, 387, 388, 390Quality assurance issues around physician performance
      441, 442, 443, 444Infection control
      492Compliance with federal, state, and local laws and regulations; physician oversight; additional duties (see also 281–282)
      501Medical direction
      520Establish and implement a relevant facility-wide quality assurance program, including a QA committee
      Tabled 1
      Scope and Severity DesignationWeight
      A – isolated event, no actual harmdeleted, not significant to our study
      B – possible pattern, no actual harm0
      C – widespread, no actual harm1
      D – isolated, no actual >minimal harm, no immediate jeopardy1
      E – possible pattern, no actual >minimal harm, no immediate jeopardy2
      F – widespread, no actual >minimal harm, no immediate jeopardy2
      G – isolated, actual harm, no immediate jeopardy3
      H – pattern, actual harm, no immediate jeopardy, substandard care10
      I – widespread, actual harm, no immediate jeopardy, substandard care10
      J – isolated, immediate jeopardy, substandard care15
      K – pattern, immediate jeopardy, substandard care20
      L – widespread, immediate jeopardy, substandard care20

      References

      1. AMDA. Certified Medical Director in Long Term Care (AMDA CMD). Available at: http://www.amda.com/certification/overview.cfm. Accessed June 17, 2009.

      2. Interpretive Guidelines for Long-Term Care Facilities. Available at: http://cms.hhs.gov/manuals/Downloads/som107ap_pp_guidelines_ltcf.pdf. Accessed June 17, 2009.

      3. NursingHome Compare. Design for Nursing Home Compare Five-Star Quality Rating System: Technical Users Guide. January 2009. Available at: http://www.cms.hhs.gov/CertificationandComplianc/13_FSQRS.asp. Accessed June 17, 2009.

      4. Obtained from the files of Cowles Research Group. Available at: http://www.longermcareinfo.com/about_oscar.html. Accessed June 17, 2009.

        • Cowles C.M.
        Nursing Home Statistical Yearbook: 2007.
        Cowles Research Group, McMinnville, OR2008 (pp.70, 72–73)
        • American Medical Directors Association
        Roles and responsibilities of the medical director in the nursing home: position statement A03.
        J Am Med Dir Assoc. 2005; 6: 411-412
        • Schnelle J.F.
        Total quality management and the medical director.
        Clin Geriatr Med. 1995; 11: 433-448
        • Schnelle J.F.
        • Ouslander J.G.
        CMS guidelines and improving continence care in nursing homes: The role of the medical director.
        J Am Med Dir Assoc. 2006; 7: 131-132
        • Zimmer J.G.
        • Watson N.M.
        • Levenson S.A.
        Nursing home medical directors: Ideals and realities.
        J Am Geriatr Soc. 1993; 41: 127-130
        • Smith R.L.
        • Osterweil D.
        The medical director in hospital-based transitional care units.
        Clin Geriatr Med. 1995; 11: 373-389
        • Colon-Emeric C.S.
        • Casebeer L.
        • Saag K.
        • et al.
        Barriers to providing osteoporosis care in skilled nursing facilities: perceptions of medical directors and directors of nursing.
        J Am Med Dir Assoc. 2005; 6: S61-S66
        • Richards Jr., C.L.
        Preventing antimicrobial-resistant bacterial infections among older adults in long-term care facilities.
        J Am Med Dir Assoc. 2005; 6: 144-151
        • Munir J.
        • Wright R.J.
        • Carr D.B.
        A quality improvement study on calcium and vitamin D supplementation in long-term care.
        J Am Med Dir Assoc. 2006; 7: 305-309
        • Pattee J.J.
        • Otteson O.J.
        Medical Direction in the Nursing Home: Principles and Concepts for Physician Administrators.
        North Ridge Press, Minneapolis, MN1991
      5. Levenson S.A. Medical Direction in Long-Term Care: A Guidebook for the Future. 2nd ed. Carolina Academic Press, Durham, NC1993
      6. Katz P.R. Calkins E. Principles and Practice of Nursing Home Care. Springer Publishing, New York1989
        • Ouslander J.G.
        • Osterweil D.
        • Morley J.
        Medical Care in the Nursing Home.
        2nd ed. McGraw-Hill, New York1997
        • Boyce B.F.
        • Bob H.
        • Levenson S.A.
        The preliminary impact of Maryland's medical director and attending physician regulations.
        J Am Med Dir Assoc. 2003; 4: 157-163
        • Caprio T.V.
        • Karuza J.
        • Katz P.R.
        Profile of physicians in the nursing home: Time perception and barriers to optimal medical practice.
        J Am Med Dir Assoc. 2009; 10: 93-97