Abstract
Background/Aims Life expectancy gains among HIV-infected patients have increased emphasis on quality multidisciplinary care. Among HIV-infected patients in Kaiser Permanente (KP), we evaluated new HIV quality of care measures that facilitate preventive services assessment. Services considered were CD4 and viral load (VL) testing, lipid screening, blood pressure (BP) management, and vaccines (influenza, pneumococcal, Hepatitis A/B, and tetanus).
Methods We included adult HIV-infected patients (median age 41; IQR 35–49) from four KP regions serving Northern and Southern California, Hawaii, and Oregon, employing KP administrative and electronic medical record (EMR) data. Eligible patients received an HIV diagnosis before 12/31/2009 and had 2+ years of utilization/membership during 1/1/2003–12/31/2010. We calculated two quality indices per service. The Prevention Index (PrI) is the proportion of patient time covered by receipt of a recommended preventive service, relative to the time during which patients were eligible to receive that service. The “target period” is the interval when service coverage is evaluated (e.g., calendar year). During the prior “observation period” of one recommended service interval (e.g., for BP, 12 months pre-coverage period), data are gathered on prior service delivery that inform the target period. The Disease Management Index (DMI) measures deviation of disease indicators from a management goal (e.g., undetectable VL) during the coverage period. Yes/no vaccine indices are based on whether the service was ever delivered.
Results 15,950 patients (90% male; 57% white) were analyzed. DMIs indicated mean patient-level control of CD4 and VL counts of >85% across years. Related PrIs indicated approximately 60% of eligible patient time across years was covered by receipt of tests for CD4 and VL. BP indicated similar control and apparent upward trends in BP exam delivery. Flu vaccinations were received by 60%–70% of patients annually and a similar proportion received at least one tetanus vaccination over the study period. Completed vaccination sequences were much lower for pneumococcal (3% of patients), hepatitis B (6%), and hepatitis A (14%), but these estimates are influenced by relatively short periods used to define completion.
Conclusions PrI and DMI represent innovative uses of EMR data for measuring HIV care quality. Future research should extend the methods to other healthcare services.
