Generalists and oncologists show similar care practices and outcomes for hospitalized late-stage cancer patients

OBJECTIVE. The objective of this work was to identify similarities and diff erences in primary attending physicians' (generalists' versus oncologists') care practices and outcomes for seriously ill hospitalized patients with m alignancy. DESIGN. This was a prospective cohort study (SUPPORT project). SETTING. Subjects were recruited from 5 US teaching hospitals; data were ga thered from 1989 to 1994. SUBJECTS. Included in the study was a matched sample of 642 hospitalized pa tients receiving care for non-small-cell lung cancer, colon cancer metastas ized to the liver, or multiorgan system failure associated with malignancy with either a generalist or an oncologist as the primary attending physicia n. MEASUREMENTS. Care practices and patient outcomes were determined from hosp ital records. Length of survival was identified with the National Death Ind ex. Physicians' perceptions of patient's prognosis, preference for cardiopu lmonary resuscitation (CPR), and length of relation-ship were assessed by i nterview. A propensity score for receiving care from an oncologist was cons tructed. After propensity-based matching of patients, practices and outcome s of oncologists' and generalists' patients were assessed through group com parison techniques. RESULTS. Generalist and oncologist attendings showed comparable care practi ces, including the number of therapeutic interventions, eg, "rescue care" a nd chemotherapy, and the number of care topics discussed with patients/fami lies. Length of stay, discharge to supportive care, readmission, total hosp ital costs, and survival rates were similar. For both physician groups, per ception of patients' wish for CPR was associated with rescue care (P < 0.03 ), and such care was related to higher hospital costs (P < 0.000). Poorer p rognostic estimates predicted aggressiveness-of-care discussions by both ty pes of physicians. Length of the patient-doctor relationship was associated with oncologists' care practices. More documented discussion about aggress iveness of care was related to higher hospital costs and shorter survival f or patients in both physician groups (P < 0.001). CONCLUSIONS. Generalists and oncologists showed similar care practices and outcomes for comparable hospitalized late-stage cancer patients. Physicians ' perceptions about patients' preferences for CPR and prognosis influenced decision making and outcomes for patients in both physician groups. Length of relationship with patients was associated only with oncologists' care pr actices. Rescue care increased hospital costs but had no effect on patient survival. Future studies should compare physicians' palliative care as well as acute-care practices in both inpatient and ambulatory care settings. Pa tients' end-of-life quality and interchange between physician groups should also be documented and compared.