August 2005

Dear Readers: Here's the Article for this Month:

Name of article:

Predicting survival in terminal cancer patients: clinical observation or quality of life evaluation?

This is a prospective cohort study carried out in 58 Italian Palliative Care (PC) Units with the aim to compare the prognostic power of patients’ quality of life (QoL) evaluations and of some clinical symptoms or problems evaluated by the physician and to assess whether their combined use could improve their predictive accuracy.

Of the 601 randomly selected terminal cancer patients, 574 were followed until death and received a prospective evaluation: clinical base evaluation, a weekly QoL assessment and a final evaluation regarding the last week of life by STAS (McCarthy and Higginson, Palliative Medicine 1991; 5: 215).

The clinical and QoL baseline data and the clinical condition during the week before death were analysed.

Performance impairments were measured by Katz’s Activities of Daily Living (ADL) (Katz and Akpom, Int J Health Serv 1976; 6: 493) and QoL by means of Therapy Impact Questionnaire (TIQ) (Tamburini et al. Ann Oncol 1992; 3: 565). TIQ is a multidimensional 36-item questionnaire filled in by the patient or proxy with respect to the previous week.

Two primary scales were derived from TIQ items: 1. physical symptoms index (PSI) (the physical well-being, fatigue, gastrointestinal symptoms); 2. the therapy impact index (TII) (functional, emotional, cognitive status, and the social interaction).

Physicians’ or nurses’ QoL evaluation was used in 28.6% of cases to replace missing data that was related to the worsening of the patient’s conditions.

Dysphagia for solids, severe dyspnea, delirium, dehydration, cachexia, hepatic and acute renal failure, oliguria, and Katz’s ADL were the clinical variables most closely associated with survival (hazard ratios (HRs) > 2.1). Only dyspnea, cachexia, oliguria and Katz’s ADL kept their strength once introduced in the Cox model. The TIQ scales most strongly associated with survival were physical well-being, fatigue, functional status and cognitive status (HRs  ranging from 1.42 to 1.71), but only fatigue showed a strong independent prognostic relevance (90% of selection). In the TIQ scales, PSI was strongly associated with survival with respect to TII (HR 1.71 vs HR 1.47).

This study shows that clinical variables are better predictors of survival than QoL in terminal cancer patients.

Why I chose this article

Many studies were carried out in order to evaluate prognostic factors of survival of cancer patients. The aim of some of these studies was to enable the study of these factors and enable identification of patients as possible candidates for MEDICARE.

As a matter of fact, in oncology and in a palliative care setting, prognostic indicators of survival are above all intended to improve patient care. They are also used to personalize and plan specific care strategies such as the best setting for the patient’s care (PC Unit, hospice, home care, nursing home), to facilitate most clinical, emotional, spiritual and ethical goals of PC and also to give patients the time to fulfil their commitments.

This is the first study comparing the prognostic power of clinical variables and other QoL indicators such as performance status, global health status, and cognitive impairment. The study shows that clinical variables are better predictors of survival than QoL in terminal cancer patients. In a future issue, I will present the following  article “Prognostic factors in advanced cancer patients. Evidence-based clinical recommendations” written by Maltoni M, Caraceni A, Brunelli C et al., in press in J Clinical Oncology.

Carla Ripamonti, MD
Member of the Board of Directors, IAHPC