Predictive effect of depression, pain and poor cognitive function on medical outcomes and quality of life in elderly cancer patients

Abstract

Context. Home health care has become a critical component of the health care system, providing nursing care for older adults in the home setting.;Objective. To determine the effect of depression, pain and poor cognitive function at baseline on medical outcomes and quality of life at one year in elderly cancer patients receiving home health care services.;Design, setting and patients. Survey of medical/surgical patients from Visiting Nurse agency in Westchester County, N.Y. Patients randomly sampled weekly from 1997-1999 from new admissions, age 65 and older. Eligibility included ability to give informed consent and to speak and understand English or Spanish with an absence of significant communication problems. Patients interviewed at home. Medical records, VNS records, and informants provided additional information. Subset of patients selected for this analysis (N=75) were treated for cancer within one year prior to receipt of VNS services. Patients were 65-91 years old (mean 76.4, sd=6.14), 66.7% female, and 14.7% minority.;Main outcome measures. Major depression diagnosed using DSM-IV etiologic criteria by "best estimate" procedures where a geriatric psychiatrist, medical internist and psychologist reviewed structured clinical patient interviews, informant data, and medical charts.;Results. In linear regression, "a great deal" of pain at baseline significantly predicted fewer IADL limitations at one year (p=.006), controlling metastatic status and IADLs at baseline. Neither depression nor cognitive function was predictive of any outcome variable.;Conclusions. The finding that baseline pain predicted fewer IADL limitations at one year may indicate that patients who reported "a great deal" of pain at baseline had an acute medical condition that was affecting IADL limitations, and that as the condition resolved, so did the pain, resulting in increased ability to perform IADLs. Alternatively, it may be that patients were successfully treated for their pain after the baseline assessment. The overall lack of significant findings may have resulted from limitations in the study which precluded the emergence of the expected relationships between predictor and outcome variables.