The incidence of attempted CPR in nursing homes.
We studied the frequency with which cardiopulmonary resuscitation (CPR) is attempted on residents of American nursing homes. Each author (all members of the Clinical Practice Committee of the American Geriatrics Society) completed a questionnaire in 1989 about policy and practice regarding CPR during 1988 in each of three to seven nursing homes, by questioning the medical or nursing director or the administrator. Because of the vagaries of nursing home record-keeping, data from some homes were allowed when they were "accurate to within 10%." Data from 58 nursing homes, totalling 10,836 bed-years were available. In 33 of these homes, accounting for 5,425 bed-years, CPR was never attempted. CPR was more likely to be foregone in nursing homes with religious affiliation than in nursing homes without (13 of 17 vs 18 of 38; chi 2 = 4.0; P less than 0.05). Religious affiliation was unknown for three nursing homes. Academic affiliation (10 of 16 vs 20 of 37 in non-affiliated nursing homes) and non-profit status (14 of 19 vs 16 of 23 in for-profit nursing homes) did not significantly affect the likelihood that CPR would never be used. In 31 of 54 nursing homes with explicit do not resuscitate (DNR) policies, CPR was never performed, compared to 2 of 4 homes without such policies. For nursing homes with complete data, there were 1,196 deaths in 32 facilities where CPR was never attempted compared to 1,294 deaths for 24 nursing homes with CPR. For 22 nursing homes without CPR, there were 2,172 emergency room transfers compared to 1,363 emergency room transfers in 18 nursing homes where CPR was attempted.(ABSTRACT TRUNCATED AT 250 WORDS)