Control groups historical

Each article was assessed for the type of evaluation and categorized (Table 1). Two factors were considered in determining the type of evaluation the presence of two or more alternatives, and the consideration of both input (costs) and outcomes. Evaluations that included two or more alternatives (i.e., concurrent control group, historical control, preintervention and postintervention design) were considered true analyses, whereas those that did not include a comparison were labeled descriptions. A description of the type of analysis was assigned to the evaluation and included the options of cost or outcome description, cost or outcome analysis, cost and outcome description, and true clinical economic evaluation. Those articles considered true clinical economic evaluations were subcategorized by type, options including cost-minimization analysis, cost-benefit analysis, cost-effectiveness analysis, and cost-utility analysis. 

Lifetime exposure of rats to hexachloroethane resulted in renal carcinomas and adenomas in Fischer-344 male rats (NTP 1989). The incidence of adenomas was 1/50 for the controls, 2/50 for animals at a dose of 10 mg/kg/day, and 4/50 for animals at a dose of 20 mg/kg/day. In the animals from the high-dose group, there were also 3/50 renal carcinomas. The number of tumors (carcinomas and adenomas) was significantly greater in exposed rats than in both controls and historical controls using the Fisher Exact Test (NTP 1989). No tumors were seen in the female rats. 

The data tend to be imprecisely gathered (often because researchers are unsure of what they are looking for), and therefore possess extreme within-group variability. Control and historical data are not used to adjust for variability or modify test performance. 

Irrespective of the specific protocols used, all carcinogenicity studies end with a statistical comparison of tumor proportions between treated and control groups. This analysis is necessary because the control incidence of most tumor types is rarely zero. In the unlikely case that a type of tumor is found in treated animals but not in concurrent or appropriate historical controls, it is reasonable to conclude that the tumor is drug-related without statistical analysis. 

The test article concentration is normally the highest nonirritating concentration. Several concentrations could be tested at the same time should one wish to establish a dose-response curve for induction. The test is easiest to perform if the vehicle is a standard nonirritating organic, such as acetone, ethanol, or dimethylformamide, or a solvent-olive oil blend. Until a laboratory develops its own historical control base, it is also preferable to include a positive control group. Either 0.25% dinitricholoro-benzene or 0.05% oxazalone are recommended for positive controls. If the vehicle for the positive control is different than the vehicle for the test material, then two vehicle control groups may be necessary. 

In a 2-year dietary study, female rats in the high dose group (1.02 mg/kg/day) had a 40% mortality rate during the last week of the study compared with 12% in controls (Hayes 1985). While the mortality rate in the control group was unusually low, the 40% mortality rate in the high dose female rats was also increased when compared with historical controls, in which the mortality rate ranged from 18 to 34%. No increase in the mortality rate of male rats was observed. Furthermore, no increase in the mortality was reported for mice exposed to 2.13 mg/kg/day (males) or 2.53 mg/kg/day (females) disulfoton in the diet for 23 months (Hayes 1983). These results support the conclusion that rats are more sensitive than mice and that female rats are more sensitive than male rats to the lethal effects of disulfoton. 

The most important comparison is with concurrent control groups. However, there are occasions when it is necessary to use historical data, that is, information from control animals of the same strain on other studies. This is more relevant if the studies were conducted in the same laboratory under similar conditions and at the same time. The incidence of a particular neoplasm is often different between laboratories and may change with time. Historical data are most useful to get an idea of the variation in the background range of frequency and also to ascertain that rare tumour t)rpes can occur spontaneously. 

The principle behind establishing a control group as opposed to a control treatment is the selection of a population as similar as possible to the group receiving the medicine under investigation. Whenever possible, a prospective rather than historical control should be used. 

Since teratology studies in mice are far less frequent than in the rat, the amount of recent historical control data may be limited. If little or no recent reference data are available, the size of the concurrent control group should be increased. 

Laboratories should compile their concurrent control group results from reproductive and developmental toxicity studies to serve as historical control data, which is an essential tool for interpreting concurrent control data and apparent changes in treatment groups. Historical control data from other laboratories and meta-databases, such as the MARTA historical control database, can also be useful tools for investigators to appreciate the broader variability in controls, and can be used, albeit with caution, to supplement their... 

Similarly, external or historical controls (groups of subjects external to the study either in a different setting or previously treated) cannot provide definitive evidence. Byar (1980) provides an extensive discussion on these issues. 

The standard deviations referred to above often provide the biggest challenge. The information for this will come from previous data for that same endpoint, from a similar population/sample of patients, treated for the same period of time etc., similarly for the success/event rate in the control group for binary data. We should try and match as closely as possible the conditions of the historical data with those pertaining to the trial being planned. 

E. Therapeutic response Thrombin-dependent tests show dose dependency [aPTT rise proportionally to dose of Refludan]. The key criteria of efficacy in two pivotal clinical trials from a laboratory standpoint were platelet recovery and effective anticoagulation. Seven days after the start of treatment with Refludan in patients with HIT, the cumulative risk of death, limb amputation, or new thromboembolic complication was substantially lower than in a historical control group. 

---