Impact of physician telephone management of nursing home residents before the initial assessment: a pilot study.

OBJECTIVE To describe physician telephone management of newly admitted nursing home residents before direct evaluation by the physician, and the effect on resident outcomes. DESIGN Retrospective chart review of 111 consecutive discharge records from two proprietary community nursing homes in Baltimore, Maryland in 1999. MEASUREMENTS Data regarding the admission process were collected, with an emphasis on physician telephone orders at admission and all subsequent telephone orders before the first physician visit. Physicians were categorized as attending physicians or on-call physicians. Unexpected outcomes defined as an unplanned admission to an acute hospital or an unanticipated death within 14 days of admission to the nursing home were identified. The relationships among resident, physician, and admission characteristics and unexpected outcomes were analyzed. RESULTS Most residents (97 of 111 (87%)) were admitted from an acute hospital, and the remaining 13% were admitted from home or another nursing home. An attending physician confirmed admission orders for 87 of 111 (78%) residents, and an on-call physician confirmed admission orders for the remainder. Physicians changed medications at the time of admission, as compared with preadmission medications, in 58 of 111 (52%) residents and ordered laboratory studies or radiographs in 59 of 111 (53%). On-call physicians were just as likely to make both types of changes as attending physicians. In the time interval after the initial telephone contact but before the first physician visit, medication changes were made in 35 of 111 (32%) residents and testing was ordered in 16 of 111 (14%). Nineteen of 111 (17%) residents were either readmitted to the hospital or died within 14 days of admission to the nursing home. These unexpected outcomes were statistically less likely to occur in the group of residents for whom physicians made medication changes at the time of admission as compared to the group for whom no medication changes were made [6 of 58 (10%) versus 13 of 53 (25%), P = 0.04, respectively], and in the group for whom tests were ordered at the time of admission as compared to not ordered [4 of 59 (7%) versus 15 of 51 (29%), P = 0.002, respectively]. There were no differences in the likelihood of unexpected outcomes when physicians made medication changes or ordered tests after the time of admission but before the first physician visit. CONCLUSIONS In this study, physicians made adjustments in medications and ordered tests for newly admitted nursing home patients before seeing the resident in the majority of cases. Unexpected outcomes including readmission to the hospital or death within 14 days of admission were less common among those residents when such changes were made at the time of admission. Further studies are needed to identify those changes as well as those resident and physician characteristics that might lead to improved outcomes.