Cancer Trial Participants Not Generalizable to Real World


Am Fam Physician. 2006 Sep 15;74(6):1023-1024.

Clinical Question: Are cancer trial participants representative of cancer patients in the real world?

Setting: Outpatient (specialty)

Study Design: Cohort (retrospective)

Synopsis: Because fewer than 5 percent of patients with cancer participate in clinical trials, the generalizability of the data to all cancer patients is potentially problematic. To evaluate this, the authors used data from a cancer registry that included more than 60,000 patients who were diagnosed between 1990 and 1997. To be included, they had to have undergone at least seven years of follow-up. The authors determined whether the patients participated in a therapeutic clinical trial within four months of diagnosis (early participants), more than four months after diagnosis (late participants), or not at all (nonparticipants). The primary outcome was survival status by the end of 2001. The researchers also assessed the presence of comorbidity, a number of markers of disease severity, and demographic variables to determine their interrelationship with survival.

After excluding patients with multiple cancers and those who had already started treatment, more than 19,000 patients were eligible for this study (5,122 early participants, 1,199 late participants, 13,019 non-participants). Participants were younger, less likely to have chronic comorbid conditions, and were more functional than nonparticipants. However, participants tended to have more distant metastasis and lymph node involvement. After taking the above-mentioned confounding factors into account, participants with localized disease had significantly shorter survival than nonparticipants. Late participants fared even worse. Participants with metastatic disease had significantly longer survival, which could affect the overall results.

Bottom Line: Patients participating in cancer trials are generally younger and healthier than nonparticipants. However, their survival rates are not necessarily better. This study is one of several that question the applicability of clinical trial results to real-world patients. (Level of Evidence: 2b)

Study Reference:

Elting LS, et al. Generalizability of cancer clinical trial results: prognostic differences between participants and nonparticipants. Cancer. June 1, 2006;106:2452–8.

Used with permission from Barry H. Cancer trials not generalizable to real world. Accessed July 3, 2006, at: http://www.InfoPOEMs.com



Want to use this article elsewhere? Get Permissions


Feb 15, 2017

Access the latest issue of American Family Physician

Read the Issue

Email Alerts

Don't miss a single issue. Sign up for the free AFP email table of contents.

Sign Up Now

Navigate this Article