Identifying prognostic factors for survival in advanced cancer patients: a prospective study.

OBJECTIVE To identify potential prognostic factors affecting the survival in patients with advanced cancer in a local palliative care unit. DESIGN Prospective cohort study. SETTING Palliative Care Unit of a regional hospital in Hong Kong. PATIENTS All advanced cancer in-patients and out-patients who were enrolled into the palliative care service of the United Christian Hospital between January and December 2002 were recruited. MAIN OUTCOME MEASURES Potential prognostic factors including demographic data, tumour characteristics, blood parameters, functional status, co-morbidities, total symptom score, and psychosocial parameters were recorded upon enrollment. RESULTS A total of 170 patients were eligible for analysis; their mean age was 69 (standard deviation, 12) years, of which 106 (62%) were male. Overall median survival was 77 (interquartile range, 31-160) days. The most frequent primary malignancy was lung (n=58, 34%), followed by liver (n= 24, 14%) and lower gastro-intestinal tract (n=24, 14%). By univariate analysis, 11 factors affected survival, including: age (P=0.040), number of metastatic sites involved (P=0.001), peritoneal metastases (P=0.009), skin metastases (P=0.011), tachycardia (P=0.009), serum albumin concentration (P<0.001), white cell count (P=0.002), Karnofsky Performance Status score (P<0.001), Hamilton Depression Scale score (P=0.004), Edmonton Symptom Assessment System score (P=0.003), and McGill Quality of Life (Hong Kong)-single item score (P=0.002). Multivariable Cox regression analysis revealed that only age (hazard ratio=0.84; 95% confidence interval, 0.73-0.96), number of metastatic sites involved (1.33; 1.13-1.56), serum albumin concentration (0.95; 0.92-0.98), Karnofsky Performance Status score (0.86; 0.78-0.96), and Edmonton Symptom Assessment System score (1.22; 1.05-1.41) were independent prognosticators. CONCLUSION Age, number of involved metastatic sites, serum albumin, Karnofsky Performance Scale score, and Edmonton Symptom Assessment System score were independent prognosticators. Further studies are needed to provide a prognostic instrument applicable in local clinical settings.