Search

We used existing data on hepatitis C prevalence, injection-related hepatitis C transmission and needle use in prisons and new data on infectiousness, to estimate the size of study required to detect injection-related hepatitis C in UK prisons. A pilot study of 500 prisoners followed for 10 weeks would have a 65% chance of detecting a hepatitis C seroconversion, conservatively assuming one injection per prisoner per week, and a 3% transmission rate per injection, but uncertainty might persist as to whether transmission had occurred during a short incarceration or before it. If the actual transmission rate was 10%, as recently documented, then such a study would have more adequate statistical power. A definitive study of 3000 prisoners for 10 weeks would expect to detect about six seroconversions, even with conservative estimates of injection frequency and transmission rate. Adequate design and power of these studies is important because of the complacency that could result from false-negative findings. We suggest six risk-factor themes that studies should document.

The Lancet now incorporates statistical review of submitted papers which remain candidates for publication after conventional review. We summarise here criticisms noted by the statistical reviewers for 191 such papers received between November, 1990, and June, 1991. Only 54% of papers were deemed acceptable or acceptable after revision; the others were either recommended for rejection (14%) or for more substantial revision and re-review (32%). Descriptions of methods and of results were found inadequate in about half of the papers; about one-quarter of papers had inadequate abstracts and conclusions. Major errors of inference were made in 48 papers and went hand in hand with major criticisms of analysis or design in those papers. The natural focus of statistical review is whether conclusions drawn are justified by study design and statistical analysis. In this, there is room for improvement by authors.

Sixty-two reports that appeared as Papers and Originals (excluding short reports) in 13 consecutive issues of the British Medical journal included statistical analysis. Thirty-two had statistical errors of one kind or another; in 18 fairly serious faults were discovered. The summaries of five reports made some claim that was unsupportable on re-examination of the data. Medical investigators should consult with people who have a real understanding of statistical methods throughout their projects.