Nursing home expenditures and improvement in reported quality scores
Hari Sharma & R. Tamara Konetzka
The University of Chicago
Objectives: Nursing homes are assumed to have an incentive to maximize profits by pursuing the least costly path to improvement in their quality, although the strength of this assumption can vary by profit status. Non-profit facilities may forgo profits and increase expenditures to improve quality of care because they cannot share profits. Reported nursing home quality, on average, has been improving in recent years; however, it is unclear whether these improvements are due to increased expenditures on patient care, greater efficiency, and/or a manipulation of the reporting system by nursing homes. In this paper, we assess whether improvements in reported qualityscores are predicted by increases in nursing home expenditures, reflecting potential investment in quality improvement.
Data/Methods: We merged data on costs obtained from Medicare’s Healthcare Cost Report Information System, facility characteristics obtained from Online Survey Certification and Reporting System, and deficiency, MDS quality, RN, and total staffing scores used by CMS for nursing home 5-star ratings from 2007-2010. Average costs per resident day and qualityscores were descriptively analyzed by baseline quality levels in2008. The relationship between changes in costs and quality scores was analyzed using facility fixed-effects regressions controlling for year and case-mix. Standard errors were clustered at the facility level. In addition, we estimated the percent contribution of changes in costs to the changes in qualityscores using predictions from the regressions. Subgroup analyses included groups stratified by levels of baseline quality and profit status.
Results: In descriptive analysis, quality, RN, and total staffing scores over time improved for most levels of baseline qualitywhile deficiency scores improved only for low-rated facilities. Regression analyses showed that improvements in deficiency, MDS quality, and to some extent in RN staffing scores were predicted by increases in costs. Subgroup analyses revealed differences in cost-score relationship for groups stratified byprofit status and baseline quality. Although changes in costs had a statistically significant relationship with deficiency and MDS quality scores across different groups of facilities, changes in costs had a meaningful contribution to changes in deficiency, MDS quality, and RN staffing scores only for highly rated facilities and non-profit facilities. The contribution of changes in costs to changes in total staffing scores was either negative or negligible in most cases.
Conclusions/Implications: Although improvements in deficiency and MDS quality rating scores were predicted byincreased costs, the relationship varied by subgroup. Moreover, the contribution of costs to changes in deficiency and MDS quality scores was meaningful only for highly rated and non-profit facilities. It is possible that nonprofit facilities’ incentive to improve quality at the expense of additional profits may have played a role in this finding. Overall, and especially for staffing, the limited explanatory power of costs in predicting changes in reported nursing home quality suggests that changes in coding of self-reported data may have played a role in improved quality scores over time.