Subgroup analysis

Subgroup analysis showed that all penicillin-induced deaths occurred between 60 and 74 years, whereas cephalosporin-induced deaths occurred between 35 and 74 years of age. Significant comorbidities included ischemic heart disease or dysarrhythmia, obstructive airway disease, mastocytosis and hypogammaglobulinemia. 

Generalized efficacy of t-pa for acute stroke. Subgroup analysis of the NINDS t-pa stroke trial. Stroke. 1997 28 2119-2125. 

More recent reports conclude that early CEA after a nondisabling ischemic stroke can be performed with perioperative mortality and stroke rates comparable to those of delayed CEA. In a subgroup analysis by the North American Symptomatic Carotid Endarterectomy Trial (NASCET) investigators, 42 patients who underwent early CEA (<30 days after stroke) were compared with 58 patients who underwent delayed CEA (>30 days), and no overall difference was demonstrated in the perioperative stroke rate (4.8% vs. 5.2%). Another recent prospective randomized study of 86 patients showed no difference in either perioperative stroke (2% in both groups) or survival rates (mean 23 months follow-up) between patients randomized to early or delayed CEA. ... 

Celecoxib, however, was effective. In the subgroup of patients with moderate to severe OA pain, the combination of glucosamine and chondroitin appeared to have a moderate effect, although this subgroup analysis must be interpreted with caution. On the other hand, a European study (the GUIDE trial) that compared a prescription glucosamine product with acetaminophen and placebo in patients with knee OA reported that glucosamine performed better versus placebo than did acetaminophen.27... 

The final question asks about the validity of the conclusions drawn by the study authors. Were the conclusions based on some overall index or ratio of costs to consequences, and was the index interpreted intelligently Did the study authors provide benchmarks to aid in the interpretation of the study, and was the robustness of the conclusions discussed in light of results of the sensitivity and/or statistical analyses Was subgroup analysis undertaken where relevant Were the results compared with those of others who have investigated the same question Were the limitations of the study and the gen-eralisability of the results discussed Were other relevant factors in the decision to adopt the intervention discussed (e.g. distribution, ethics) And... 

The CHPA commissioned a study entitled Self-Care in the New Millenium American Attitudes Toward Maintaining Personal Health and Treatment. They conducted 1505 telephone interviews in January of 2001, using random telephone numbers. African-Americans and Hispanics were oversampled to conduct in-depth subgroup analysis. Of particular interest is the finding that 96% of respondents felt confident that they could take care of their own health. This might explain why so many people want access to pharmacologically active botanicals. These products do not require a prescription and thus allow patients to treat themselves. 

Vollmer WM et al Effects of diet and sodium intake on blood pressure Subgroup analysis of the DASH-Sodium trial. Ann Intern Med 2001 135 1019. [PMID 11747380]... 

A prospective placebo-controlled subgroup analysis has shown a marked decrease in the progression of intimal-medial thickness. 

W. I. Fushchych, L. F. Barannik, and A. F. Barannik, Subgroup Analysis of the Galilei and Poincare Groups and Reduction of Nonlinear Equations, Naukova Dumka, Kiev, 1991 (in Russian). 

Qotet B, Bellos N, Mohna JM, Cooper D, et al. 2007. Efficacy and safety of darunavir-ritonavir at week 48 in treatment-experienced patients with HIV-1 infection in POWER 1 and 2 A pooled subgroup analysis of data from two randomized trials. Lancet. 369 1169-1178. 

This indication was considered in three trials. The first was the PCI-CURE study (44) (n = 2658 patients), a prespecified subgroup analysis of CURE. This trial studied the benefit of pretreatment with clopidogrel (median 10 days) before PCI. At one-month follow-up, there was a significant (P = 0.04) reduction of cardiovascular death and Ml (from 4.4% to 2.9%). 

The third trial was a subgroup analysis of the CLARITY (29) trial performed in acute Ml. It was demonstrated that in STEM I patients, treated with fibrinolytic and who underwent PCI during the hospitalization period (n = 1863 patients), the dual antiplatelet treatment was able to reduce major vascular events (death, Ml, and stroke) from 12% to 7.5% (RRR = 0.59 95% Cl 0.43-0.81 P = 0.001). Thus, the treatment with clopidogrel + aspirin of 43 STEMI patients followed by PCI prevents one major vascular event. 

The study design included three comparisons ACTIVE W ACTIVE A, and ACTIVE I in 14,000 patients, (Maximum follow-up was for 48 months), The primary endpoint was the time to first vascular event (stroke, Ml, vascular death, systemic emboli). ACTIVE W arm was halted when 6600 patients were enrolled because there a clear benefit from warfarin treatment compared to clopidogrel + aspirin 3.63% of vascular events versus 5,64% (P = 0,0002). Subgroup analysis showed that these disappointing results were observed in patients on warfarin prior to study (HR = 1.5, P = 0.0006), but there was no difference between the two strategies—when the patients were not on warfarin prior to study (HR = 1.32, P = 0,17), Nevertheless, further results are awaited from the ACTIVE-A arm (ASA or ASA + clopidogrel) in patients who cannot or would not take OAC. 

In the PCI-CURE (Clopidogrel in Unstable Angina to Prevent Recurrent Events) trial, subgroup analysis of diabetic patients... 

In the CAPRIE trial (Clopidogrel vs. Aspirin in Patients at Risk of Ischemic Events), clopidogrel reduced the risk of Ml, stroke, or vascular death by 23.8% compared with aspirin in patients with PAD (48). Although this is an impressive reduction in major events, the benefits of clopidogrel over aspirin were identified as a subgroup analysis rather than a primary endpoint. 

Univariate subgroup analysis is of relatively limited value, even when done reliably, in situations where there are multiple determinants of the individual response to treatment. In this situation, targeting treatment using risk models can be useful, particularly in... 

Rothwell PM (2005b). Subgroup analysis in randomised controlled trials importance, indications and interpretation. Lancet 365 176-186... 

There are some clinically useful subgroup observations in the pooled analysis of the endarterectomy trials, but the results of univariate subgroup analysis are often of only... 

---