The prospective observational study by Penketh et al.1 has serious methodological deficiencies and should be read with caution.
In the first instance, there are serious selection issues. A total of 71 women were recruited over 5 years, approximately one per month. It is not clear from the paper how many women were excluded. It is difficult to believe that the researchers of this ‘prospective’ study expected such low recruitment.
Secondly, the absence of a comparison group is a substantial deficiency if only because of the Hawthorn effect. The report has no benefit if it just reports that ‘it can be performed’. This fact is already known, even in the outpatient setting. It missed an opportunity to compare their group with women who either underwent abdominal hysterectomy (the indications are also valid for abdominal hysterectomy) or comparison with women recruited for vaginal hysterectomy who had a planned admission, or an unplanned admission. If this study had used the latter two groups as comparison groups, it would have given valuable information about women who are not suitable for 24-hour discharge.
Among the many faults in the methods and results sections, tables 1 and 2 show a comparison between the consultant operation and the registrar’s operation. In fact, the comparison groups were ‘the consultant’ and the ‘registrar assisted by the consultant’. The tests of significance are meaningless because the groups in the Tables are not independent. We are also not given enough information to access the test of significance (quoted as P-value). Therefore, Tables 1 and 2 contribute nothing.
Finally, the main endpoints chosen in the study (duration of the operation and mean blood loss) and the results in all three tables have no relevance to the objectives in the title and introduction (safety and acceptability). Acceptability was not reported at all!
Closer reading of the whole paper makes it difficult to justify its publication in any journal! The extensive introduction and discussion sections bear little resemblance to the objectives and the results associated with these objectives.
Beware of prospective studies that have or add no value.