A quality assessment of randomized control trials of primary treatment of breast cancer.

The methodology of randomized control trials (RCTs) of the primary treatment of early breast cancer has been reviewed using a quantitative method. Sixty-three RCTs comparing various treatment modalities tested on over 34,000 patients and reported in 119 papers were evaluated according to a standardized scoring system. A percentage score was developed to assess the internal validity of a study (referring to the quality of its design and execution) and its external validity (referring to presentation of information required to determine its generalizability). An overall score was also calculated as the combination of the two. The mean overall score for the 63 RCTs was 50% (95% confidence interval [CI] = 46% to 54%) with small and nonstatistically significant differences between types of trial. The most common methodologic deficiencies encountered in these studies were related to the randomization process (only 27 of the 63 RCTs adopted a truly blinded procedure), the handling of withdrawals (only 26 RCTs included all patients in the analyses), the description of the follow-up schedule (only 12 RCTs reported adequately), the report of side effects (adequate information given in 33 RCTs), and the description of the patient population (satisfactory in 29 RCTs). Telephone calls to the principal investigators improved the quality scores by seven points on a scale of 100, indicating that some of the deficiencies lay in reporting rather than performance. There was evidence that quality has improved over time and that the increasing tendency of involving a biostatistician in the research team was positively associated with the improvement of the internal validity but not with the external.

Efficacy and safety of nonsteroidal antiinflammatory drugs for cancer pain: a meta-analysis.

PURPOSE: To assess the efficacy and safety of nonsteroidal antiinflammatory drugs (NSAIDs) in the treatment of cancer pain by meta-analyses of the published randomized control trials (RCTs). PATIENTS AND METHODS: Twenty-five studies met inclusion criteria for analysis. Of these, 13 tested a single-dose effect, nine multiple-dose effects, and three both single- and multiple-dose effects of 16 different NSAIDs in a total of 1,545 patients. Baseline pain intensity (when provided) of moderate or higher was indicated in 81% of patients. RESULTS: Single-dose NSAID studies found greater analgesic efficacy than placebo, with rough equivalence to 5 to 10 mg of intramuscular morphine. Pain scores differed insignificantly for aspirin versus three other NSAIDs. Analgesic responses to low- and high-dose NSAIDs suggested a dose-response relationship, but this was not statistically significant. Recommended and supramaximal single doses of three NSAIDs produced comparable changes in pain scores, which indicates a ceiling analgesic effect. Common side effects included upper gastrointestinal symptoms, dizziness, and drowsiness. The incidence of side effects showed a trend to increase with dose, without a ceiling effect, and to increase with multiple doses. Single or multiple doses of weak opioids (WO) alone or in combination (WO/C) with nonopioid analgesics did not produce greater analgesia than NSAIDs alone. Single doses of WO/C analgesics produced more side effects than NSAIDs alone, although both side effect incidence and patient dropout rates were equal when multiple doses were administered. CONCLUSION: These findings question whether the traditional World Health Organization (WHO) second analgesic step (addition of a weak opioid when pain is inadequately treated by a nonopioid analgesic alone) is warranted. A lack of comparable studies precluded testing the hypothesis that NSAIDs are particularly effective for malignant bone pain.

Sensitivity and specificity of clinical trials. Randomized v historical controls.

The relative accuracy of randomized control trials (RCTs) and historical control trials (HCTs) in determining effective therapies has not been compared since there is no external verification of efficacy. We reviewed six therapies studied by both methods. Most HCTs concluded therapy was better than control, but few RCTs agreed. We calculated sensitivity and specificity for each type of trial by combining published results with all possible combinations of effectiveness. The sensitivity of HCTs was 0.80 to 1.00 (mean, 0.90) and specificity was 0.0 to 0.27 (mean, 0.11). The sensitivity of RCTs was 0.0 to 0.27 (mean, 0.12) and specificity was 0.67 to 1.00 (mean, 0.88). Defects of RCTs are more easily corrected than those of HCTs. Readers should consider trial design and the probability of errors when deciding how much credence to give to a clinical trial.

Meta-analytic evidence against prophylactic use of lidocaine in acute myocardial infarction.

Although lidocaine prophylaxis reduces the incidence of ventricular fibrillation during acute myocardial infarction (AMI), randomized control trials (RCTs) have not demonstrated any significant mortality effect of this therapy. We conducted a meta-analysis of 14 RCTs of lidocaine prophylaxis during AMI to detect any mortality effect. Six prehospital- and eight hospital-phase RCTs that randomized totals of 7656 and 1407 patients, respectively, were selected and reviewed in a blinded fashion. Mortality data were evaluated according to therapy type, reporting interval, and patient category. The prehospital-phase RCTs showed no meaningful mortality effect (risk difference, 0.0184; 95% confidence interval, -0.048 to +0.012). The hospital-phase RCTs showed a statistically significant increase in mortality during the treatment period for lidocaine recipients (risk difference, 0.029; 95% confidence interval, +0.004 to +0.055). These results confirm that lidocaine administered to monitored patients during the prehospital phase of AMI will not reduce mortality by a clinically important amount and suggest that lidocaine administered in the hospital phase of monitored, uncomplicated AMI may increase mortality among recipients with proved AMI.

Reporting of age data in clinical trials of arthritis. Deficiencies and solutions.

OBJECTIVE: We studied recent nonsteroidal anti-inflammatory drug (NSAID) randomized control trials of arthritis to identify the age and number of older people (> or = 65 years) and to document the way information on age was presented. We hypothesized that older people, who are most likely to take NSAIDs are underrepresented and underreported. STUDY SELECTION AND DATA EXTRACTION: All NSAID articles (n = 1008) in MEDLINE between September 1987 and May 1990 were identified. Eighty-three trials employing NSAIDs in a randomized control trial of arthritis reported in 73 articles were identified and studied in detail for age-related information. RESULTS: A total of 9664 subjects with a female-to-male ratio of 2.3:1 were enrolled. Forty-four trials studied osteoarthritis (53.0%), 37 studied rheumatoid arthritis (44.6%), and two studied both conditions (2.4%). More than half of the studies reviewed included people 65 years of age or older, only 207 people in this older age group could be identified (2.1%). While there was inclusion of the 'young-old' (65 to 74 years of age), only 14 of the 9664 people studied were between 75 and 84 years of age, and no one 85 years or older could be identified. The inclusion of the young-old is documented by the weighted mean age that ranged from 59.6 to 64.9 years for patients with osteoarthritis (mean, 62.9; SD, 1.67) and from 47.4 to 53.0 years (mean, 49.9; SD, 2.16) for those with rheumatoid arthritis. CONCLUSION: We demonstrate that older people, who represent a high proportion of the population treated with NSAIDs in practice, are generally omitted from drug trials. Recommendations designed to improve the reporting of age information to make clinical trials more informative and applicable to older people are presented.

A study of manufacturer-supported trials of nonsteroidal anti-inflammatory drugs in the treatment of arthritis.

BACKGROUND: To study the relation between reported drug performance in published trials and support of the trials by the manufacturer of the drug under evaluation, we studied a sample of trials of nonsteroidal anti-inflammatory drugs (NSAIDs) used in the treatment of arthritis. METHODS: All randomized control trials of NSAIDs published between September 1987 and May 1990 identified by MEDLINE were reviewed. If an article met the following criteria (n = 61), it was selected: trials involving adult patients with osteoarthritis or rheumatoid arthritis (n = 180), use of nonsalicylate NSAIDs marketed in the United States (n = 101), randomized control trial (n = 81), duration of the trial 4 or more days (n = 78), and use of an efficacy outcome measure (n = 61). Reviewers, "blinded" to manufacturer status, evaluated the narrative interpretation of results and extracted numeric data on efficacy and toxicity. Manufacturer-associated trials were defined as those that acknowledged an association with a pharmaceutical manufacturer. Because of the scarcity of non-manufacturer-associated trials (n = 9), we report only on the manufacturer-associated articles. RESULTS: Fifty-two publications (85.2%) representing 56 trials were associated with a manufacturer. The manufacturer-associated drug was reported as comparable with (71.4%) or superior to (28.6%) the comparison drug in all 56 trials. These narrative claims of superiority were usually justified with trial data. Of the trials identifying one drug as less toxic (n = 22), the manufacturer-associated drug's safety was reported as superior to the comparison drug in 86.4% of cases. Justification for the narrative interpretation of the trial findings regarding less toxicity was provided in only 12 (54.5%) of 22 trials. CONCLUSION: The manufacturer-associated NSAID is almost always reported as being equal or superior in efficacy and toxicity to the comparison drug. These claims of superiority, especially in regard to side effect profiles, are often not supported by trial data. These data raise concerns about selective publication or biased interpretation of results in manufacturer-associated trials.

Safety of nifedipine in patients with hypertension: a meta-analysis.

Our objective was to compare cardiovascular event rates in patients with mild or moderate hypertension who received nifedipine with active drug controls. We performed a MEDLARS search using the MeSH heading "hypertension" and the text word "nifedipine" to identify all articles that were published between 1966 and August 1995 in English, French, German, Italian, and Spanish languages and that involved human subjects. The computerized search was supplemented by a manual search of article bibliographies. Review of 1880 citations revealed 98 randomized controlled clinical trials that met protocol criteria. Articles were extracted independently by two doctors who were blinded for author, institution, and treatment regimen, using a structured, pretested extraction form. Differences of opinion were resolved by consensus. Fourteen events occurred in 5198 exposures (0.27%) to nifedipine and 24 events in 5402 exposures (0.44%) to other active drug controls. Unadjusted odds ratios for nifedipine versus controls were 0.49 (95% confidence interval [CI], 0.22-1.09) for definitive events (death, nonfatal myocardial infarction or stroke, revascularization procedure) and 0.61 (95% CI, 0.31-1.17) for all events (definitive plus increased angina). The odds ratio for nifedipine monotherapy (sustained- or extended-release in 91% of exposures) was nonsignificantly higher for definitive and all events (odds ratio, 1.40; 95% CI, 0.49-4.03 and odds ratio, 1.39; 95% CI, 0.59-3.32, respectively). The odds ratio for nifedipine in combination with another drug was significantly lower for definitive and all events (odds ratio, 0.09; 95% CI, 0.01-0.66 and odds ratio, 0.15; 95% CI, 0.03-0.65, respectively). Differences in odds ratio for nifedipine monotherapy and combined therapy were statistically significant (P=.02 for definitive events and P=.001 for all events). Results support the safety of sustained- and extended-release nifedipine in the treatment of mild or moderate hypertension when it is used in combination with other drugs.

Safety of nifedipine in angina pectoris: a meta-analysis.

-Our objective was to compare cardiovascular event rates in patients with stable angina receiving nifedipine as monotherapy or combination therapy and in active drug controls. A MEDLARS search of published articles from 1966 to 1995 in English, French, German, Italian, or Spanish, supplemented by a manual search of bibliographies, identified 60 randomized controlled trials that met protocol criteria. Blinded articles were extracted by 2 physicians. The pooled risks of death, withdrawal, and cardiovascular event were computed and expressed as odds ratios (ORs) for all nifedipine formulations and relative to same study control drug regimens. Thirty cardiovascular events were reported in 2635 nifedipine exposures (1.14%) and 19 events in 2655 other active drug exposures (0.72%). Unadjusted ORs for nifedipine versus controls were 1.40 (95% CI, 0.56 to 3.49) for major events (death, nonfatal myocardial infarction, stroke, revascularization procedure), 1.75 (95% CI, 0.83 to 3.67) for increased angina, and 1.61 (95% CI, 0.91 to 2.87) for all events (major events plus increased angina). Episodes of increased angina were more frequent on immediate-release nifedipine (OR, 4.19 [95% CI, 1.41 to 12.49]) and on nifedipine monotherapy (OR, 2.61 [95% CI, 1.30 to 5.26]). The OR for immediate-release nifedipine was significantly higher than that for sustained-release/extended-release nifedipine (P=0.001), and the OR for nifedipine monotherapy was higher than that for nifedipine combination therapy (P=0.03). Increased risks of cardiovascular events in patients with stable angina on nifedipine were due primarily to more episodes of increased angina, confined to the immediate-release formulation and to nifedipine monotherapy.

Dipyridamole in the treatment of angina pectoris: a meta-analysis.

A meta-analysis was performed to reevaluate the efficacy of dipyridamole for prophylaxis of angina pectoris. We found 10 articles that reported 11 randomized control trials published between 1960 and 1970. Three trials found a statistically significant benefit for the drug vs placebo, four showed a positive trend, two found no difference, and two showed a slight trend favoring placebo. When the results of all 11 trials were combined, two different statistical methods showed a statistically significant benefit from the drug. These combined results must be interpreted cautiously because of excluded patients and other methodologic variations in the studies, as well as evidence from other studies that dipyridamole may aggravate angina. Nevertheless, we conclude that there is some evidence for efficacy of the drug and believe the question should be restudied in larger and better-designed trials.

Effects of ascorbic acid on the common cold. An evaluation of the evidence.

Of 14 clinical trials of ascorbic acid in the prevention and treatment of the common cold, the data from 8 were considered well enough gathered to be creditable and to warrant combining for an over-all assessment of efficacy. Differences in mean prorated numbers of colds per year and durations of illness were 0.09 plus or minus 0.06 (plus or minus 1 standard error) and 0.11 plus or minus 0.24, respectively, favoring ascorbic acid over the placebo. These are minor and insignificant differences, but in most studies the severity of symptoms was significantly worse in the patients who received the placebo. In one study lasting 9 months, a large number of the volunteers tasted their capsules and correctly guessed what group they were in. All differences in severity and duration were eliminated by analyzing only the data from those who did not know which drug they were taking. Since there are no data on the long-term toxicity of ascorbic acid when given in doses of 1 g or more per day, it is concluded that the minor benefits of questionable validity are not worth the potential risk, no matter how small that might be.

Randomized versus historical controls for clinical trials.

To compare the use of randomized controls (RCTs) and historical controls (HCTs) for clinical trials, we searched the literature for therapies studied by both methods. We found six therapies for which 50 RCTs and 56 HCTs were reported. Forty-four of 56 HCTs (79 percent) found the therapy better than the control regimen, but only 10 of 50 RCTs (20 percent) agreed. For each therapy, the treated patients in RCTs and HCTs of the same therapy was largely due to differences in outcome for the control groups, with HCT control patients generally doing worse than the RCT control groups. Adjustment of the outcomes of the HCTs for prognostic factors, when possible, did not appreciably change the results. The data suggest that biases in patient selection may irretrievably weight the outcome of HCts in favor of new therapies. RCTs may miss clinically important benefits because of inadequate attention to sample size. The predictive value of each might be improved by reconsidering the use of p less than 0.05 as the significance level for all types of clinical trials, and by the use of confidence intervals around estimates of treatment effects.

Diabetic treatment and primary ventricular fibrillation in acute myocardial infarction.

The relation between mode of therapy and mortality rate and incidence of primary ventricular fibrillation was studied in 265 patients with diabetes mellitus and acute myocardial infarction. Sixty patients were being treated with diet only, 54 were receiving insulin and 151 were taking oral hypoglycemic agents. Fourteen patients (5.3 percent) had primary ventricular fibrillation, and all but one died. No statistically significant association was found between the incidence of primary ventricular fibrillation and the type of treatment for diabetes mellitus. Sixty-four (24.2 percent) of the 265 patients died during hospitalization. Mortality was greater among diabetic patients receiving oral therapy. However, after adjusting for age and sex, the difference among these three treatment regimens did not reach the P less than 0.05 level of significance.

Meta-analysis of randomized controlled trials as a method of estimating rare complications of non-steroidal anti-inflammatory drug therapy.

The design of randomized controlled trials to assess the efficacy of pharmacological measures for the prevention of the gastrointestinal side-effects of anti-inflammatory drugs requires an accurate estimate of excess risk under controlled conditions. Photocopies of 952 randomized controlled trial publications were obtained after scanning titles and abstracts of a MEDLINE computer search, 427 were excluded for obvious reasons, and 525 were again photocopied after obliterating source and results. Selection criteria were: the presence of a non-anti-inflammatory drug control group; at least 4 days of therapy; at least 3 days without anti-inflammatory drugs before randomization; no complicating background drugs; mention of side-effects; and a clear differentiation of gastrointestinal complications. Observer error, with two independent readings, for inclusion suitability in the study was 19% for Methods and 9% for Results. For the 44 aspirin trials, the mean therapy duration was 357 days; the unweighted rate difference between therapy and control groups ( +/- 1 S.E.M.) for ulcer was 0.006 +/- 0.003, for gross haemorrhage 0.006 +/- 0.002 and for unspecified gastric symptoms 0.03 +/- 0.01. In 123 non-aspirin non-steroidal anti-inflammatory drug (NA-NSAID) trials, the mean duration was 67 days; the unweighted rate difference for ulcer was 0.0005 +/- 0.0003, for gross haemorrhage 0.007 +/- 0.004 and for unspecified gastric symptoms 0.02 +/- 0.005. Risk differences were also pooled using the DerSimonian and Laird method, which weights studies inversely according to variance. Using this method, only the unspecified gastric symptoms for non-aspirin non-steroidal anti-inflammatory drugs (NA-NSAIDs) and the haemorrhage for aspirin were found to be statistically significant. Longer studies have higher risk differences. Randomized control trials to determine prophylactic efficacy against haemorrhage (that is, to demonstrate a reduction of ulcer rate in the therapy group to the rate of controls) would require 190 patients in each group for NA-NSAIDs in studies of 2-6 months; 950 subjects would be needed to detect a 50% reduction. Randomized control trials to determine a reduction in ulcer rate to that of controls in patients on aspirin for more than 6 months would require 700 subjects in each group; 3346 subjects would be needed to detect a 50% reduction. Such studies are feasible.

Cumulative meta-analysis of clinical trials builds evidence for exemplary medical care.

Cumulative meta-analysis of clinical trials (a Bayesian interpretation for accumulating evidence) will profoundly affect medical care by summarizing evidence in the assessment of technology innovations. Application of the technique to the randomized control trials (RCTs) of streptokinase treatment of acute myocardial infarction, reduction of peri-operative mortality by antibiotic prophylaxis, and prevention of death from bleeding peptic ulcers has revealed efficacy years before it was suspected by any other means. Arrangement of the trials according to event rate in the controls, effect sizes, quality of the trials or according to covariables of interest has supplied unique information. If carried out prospectively the technique supplies invaluable information regarding indications for another trial, the number of patients necessary to determine the validity of past trends, and the type of patients who might be benefitted. Careful examination in a cumulative manner of the prior trials can reduce the need for future large trials.

Validation of a meta-analysis: the effects of fish oil in rheumatoid arthritis.

The purpose of this study was to validate the results of a meta-analysis showing the efficacy of fish oil in rheumatoid arthritis with the results of a re-analysis of the complete primary data set. A Medline search yielded seven published papers. Three additional trials were found by contacting authorities in the field. Inclusion criteria included (1) a double-blind, placebo-controlled study, (2) use of at least one of seven predetermined outcome measures, (3) results reported for both placebo and treatment groups at baseline and follow-up, (4) randomization, and (5) parallel or cross-over design. Papers were scored for quality. Demographic and outcomes variables were collected. For the re-analysis of the primary data, the same variables were abstracted for the 395 individual patients randomized. The meta-analysis demonstrated that dietary fish oil supplementation for 3 months significantly reduced tender joint count (rate difference [RD] [95% CI] = -2.9 [-3.8 to -2.1] [p = 0.001]) and morning stiffness (RD [95% CI] = -25.9 [-44.3 to -7.5] [p < 0.01]) as compared with heterogeneous dietary control oils. The re-analysis of the primary data confirmed a significant reduction in tender joint count (p = 0.001) and in morning stiffness (p < 0.02) in the parallel analysis that ignored interaction terms. The analyses that included an interaction term between site and treatment again confirmed a significant reduction in tender joint count. The results for morning stiffness were similar to the meta-analysis, but did not quite reach statistical significance (p = 0.052-0.083). The relative improvements in the other outcome variables did not reach statistical significance. Use of fish oil improved the number of tender joints and duration of morning stiffness at 3 months as analyzed by both meta- and mega-analysis. The fuller mega-analysis confirmed the results of the meta-analysis. The advantages of mega-analysis were as follows: (1) the ability to analyze the homogeneity of the patient populations, (2) the ability to make clinically sensible adjustments in the form of the comparison, and (3) the ability to examine subsets of the data.

Are randomized control trial outcomes influenced by the inclusion of a placebo group?: a systematic review of nonsteroidal antiinflammatory drug trials for arthritis treatment.

Placebo groups are often included in randomized control trials evaluating drug therapy, yet we know little about the placebo effect. The purpose of our study was to evaluate how the presence of a placebo group in a randomized control trial (RCT) influences the patients' ratings of the efficacy of an active drug therapy and their reporting of its adverse effects. We identified studies published between 1966 and 1994 using MEDLINE. Randomized control trials evaluating acetylsalicylic acid, diclofenac, or indomethacin for the treatment of osteo or rheumatoid arthritis were included in our sample. Two investigators independently extracted data. Fifty-eight treatment arms met our inclusion criteria and were available for analysis. Twenty-five treatment arms evaluated a nonsteroidal antiinflammatory drug (NSAID) in placebo control trials and 33 in comparative trials. Using a logistic regression model to adjust for the differences between the evaluated drugs and between the types of arthritis, we found that patients receiving an NSAID in a placebo control trial were more likely to withdraw due to inefficacy (OR=1.3; 95% CI, 1.0 to 1.6; P=0.04). Using a similar model, withdrawals due to adverse effects were found to be more common when the NSAID was given in trials that did not include a placebo group (OR=1.5; 95% CI, 1.1 to 1.9; P=0.002) as were reports of cutaneous (OR=4.2; 95% CI, 1.7 to 9.9), gastrointestinal (OR=1.6; 95% CI, 1.3 to 2.0), and other types (OR=5.3; 95% CI, 3.8 to 7.4) of adverse effects. Although reports of central nervous system adverse effects were more frequent in the comparative trials, this difference was not significant. Including a placebo group in a RCT changes how patients rate the efficacy and adverse effects of their therapy. Our results highlight the need to consider the placebo effect in the design and analyses of clinical trials.

Meta-analytic stimulus for changes in clinical trials.

The advent of meta-analysis, especially when performed cumulatively, raises many questions about how best to approach the conduct of clinical trials in the evaluation of new treatments. We need to be assured that bias is minimized by proper experimental procedures and that clinical data, on the whole and in subgroups, are presented so that they can be effectively combined in meta-analysis. We need to re-examine the idea that we should not start a randomized control trial unless sufficient patients are available to avoid reasonable type I and II errors. Meta-analyses will come to the rescue, provided trials continue to be published at the present rate. We need to perform meta-analyses before undertaking each additional trial, and we need to base estimates of trial size on past data as well as the expected control rates and the differences we do not want to miss. In clinical trials of new interventions attempting to disprove the null hypothesis may be inappropriate because past data so often suggest or even establish that it is not true. Furthermore we need to recognize that trends (p > 0.05) can be both clinically and statistically important, and we must abandon the notion that if p is not < 0.05, the treatment is ineffective. In performing meta-analyses we need to worry about minimizing bias and error and consider the differences between the random and fixed effects models and between reporting results as an odds ratio versus difference in risk, with the control rates given. Experiences with cumulative meta-analysis have required that we think about all of these problems.

A meta-analysis of randomized placebo control trials in Fontaine stages III and IV peripheral occlusive arterial disease.

In patients with Fontaine Stage III and IV POAD unsuitable for arterial reconstruction, Iloprost, a prostacyclin analogue, has been shown in six RCTs to have a significant (p < 0.05) beneficial effect with regards to the probability of being alive with both legs at six months follow-up. Iloprost has significant (p < 0.05) beneficial effects over placebo on ulcer healing and pain relief, but these were relatively soft endpoints to study when side effects may have unblinded many observers and patients. Further studies are indicated to investigate the possible benefit of repeated courses of treatment with Iloprost in patients with non-reconstructable Fontaine Stage III and IV POAD as well as studies looking at patients who may be suitable only for relatively high risk reconstructions. Meta-analysis of all other RCTs of pharmacotherapeutic agents in patients with Fontaine Stage III and IV POAD showed no significant benefit over placebo for any of the endpoints reported.

The Lugano statements on controlled clinical trials.

During a consensus conference in Lugano, Switzerland, 175 statements on controlled clinical trials were drafted by 47 representatives from academia, governmental registration agencies and industry in nine countries. Their opinion on these statements was similar to that of 47 'matched pairs' who did not attend the conference. Thus, the opinion of participants and non-participants appears to reflect the general opinion of those currently involved in designing, conducting and analysing controlled clinical trials. The Lugano statements give answers to the following questions: Is the controlled clinical trial in a crisis? What is the motivation to perform controlled clinical trials? Is it possible for a physician participating in a controlled clinical trial to act in the patient's best interest? Is it possible to obtain truly informed consent in a controlled clinical trial? When is it ethical to withhold active treatment in a controlled clinical trial? What are the controversial issues in the design of a good controlled clinical trial? Is there a double standard with respect to efficacy and adverse drug reactions in controlled clinical trials? What are the alternatives to controlled clinical trials and when should they be performed? How can sponsor bias be minimized? How should an ethics committee decide whether a controlled clinical trial is ethical? Should registration agencies become directly involved in the planning and conduct of controlled clinical trials? Do the declarations of Tokyo and Helsinki facilitate the conduct of ethically valid controlled clinical trials? Is it possible to create an international standard for the conduct and regulation of controlled clinical trials? Why do messages from controlled clinical trials filter into medicine so slowly? Is it possible to bridge the gap between controlled clinical trials and clinical reality? What are the costs of doing and not doing controlled clinical trials? When should drug companies decide to start a trial programme with a specific compound? Is there public hostility against controlled clinical trials? If so, how can it be reduced? The respondents almost unanimously felt that controlled clinical trials are a must: the public must be told that progress in medicine depends on controlled clinical trials, that patients often benefit from participating in them and that the alternative, practising in the face of constant uncertainty, is worse than the possible disadvantages related to the conduct of the trial.

Can meta-analysis help uncertainty in surgery for otitis media in children.

While otitis media is perhaps the most common disease of childhood that receives medical attention, there is little agreement concerning the efficacy of the medical and surgical therapies employed to try to alleviate its symptoms or hasten its natural resolution. Because various surgeries including adenoidectomy, myringotomy, and insertion of tympanostomy tubes are frequently involved in the treatment of otitis media with effusion (OME), it is likely the most expensive condition being managed in national terms. In an attempt to elucidate the most appropriate management of this condition, a meta-analysis was attempted to the 12 randomized control trials of surgical treatments for OME in children, published between 1966 and 1990. Heterogeneity both in the populations and comparisons studied and in the outcomes presented made meta-analysis an inappropriate method for clarifying this area of clinical uncertainty. Important elements in the design of randomized control trials that should be included in future studies of treatment for OME are therefore discussed.