Re: Increased mortality associated with weekend hospital admission: a case for expanded seven day services?
When the authorship group discussed the publication of our updated analyses of the risks associated with weekend admission we decided to submit an ‘Analysis’ article rather than the more conventional scientific paper for two reasons. First the updated analysis would have not been original, as we repeated an existing and published methodology[1] and simply applied it to a newer data set. While we made a small number of improvements to the methods, which we have described, these could not reasonably be considered to distance the update from our original published work. Second, and more fundamentally, we were committed to describe our somewhat complex and sophisticated analyses, and the results, in a manner which made them accessible to readers. Reading the rapid responses from a growing number of clinicians it seems that we have failed somewhat in that aim.
We took the a priori decision to include all overnight admissions since the urgency of admission carries different risks on different days. An analysis including only emergency admissions, which several respondents requested, provides answers which are remarkably similar to the overall result (including some 81% of the deaths) although it excludes some very ill patients (Figure). However the supportive analysis of the subgroup of emergency patients clearly demonstrates that this decision did not drive the results and is remarkably similar to the analysis for the emergency group published in 2012 [1]. This finding will not be surprising to those with some knowledge of epidemiology.
Several correspondents have requested that we share ‘the raw data’, and while we would welcome the possibility of doing this, for governance reasons it is simply not possible. However, these data are available from the HSCIC to those with adequate provision for data management and security. Large datasets such as the one used for our analyses (≈150m) observations require specialist computer equipment and software to setup and undertake the analyses.
It would have been tempting simply to let the trickle of responses run their course, but for the comments that they formed a kind of ‘informal peer review’, and that taken collectively they undermine the validity of our analyses. In a quite remarkable editorial in JAMA, Drummond Rennie observed that the stream of correspondence complaining about the validity a published article contained only letters, and simply receiving letters did not constitute grounds for rejecting important clinical findings.[2] None of the comments raised in the rapid responses undermine the validity of our findings. It is curious to observe correspondents making light of a large number of unexplained deaths over the weekend period instead of these fuelling a consideration on the optimisation of service provision to meet patient need.
References
1. Freemantle N, Richardson M, Wood J, Ray D, Khosla S, Shahian D, Roche WR, Stephens I, Keogh B, Pagano D. Weekend hospitalization and additional risk of death: An analysis of inpatient data. Journal of the Royal Society of Medicine, 2012; 105: 74-84.
2. Rennie D. Smoke and Letters. JAMA. 1993;270(14):1742-1743.
Competing interests:
No competing interests
28 October 2015
Nick Freemantle
Professor
Domenico Pagano
University College London
Department of Primary Care and Population Health, UCL Medical School (Royal Free Campus), London, NW3 2PF, UK
Rapid Response:
Re: Increased mortality associated with weekend hospital admission: a case for expanded seven day services?
When the authorship group discussed the publication of our updated analyses of the risks associated with weekend admission we decided to submit an ‘Analysis’ article rather than the more conventional scientific paper for two reasons. First the updated analysis would have not been original, as we repeated an existing and published methodology[1] and simply applied it to a newer data set. While we made a small number of improvements to the methods, which we have described, these could not reasonably be considered to distance the update from our original published work. Second, and more fundamentally, we were committed to describe our somewhat complex and sophisticated analyses, and the results, in a manner which made them accessible to readers. Reading the rapid responses from a growing number of clinicians it seems that we have failed somewhat in that aim.
We took the a priori decision to include all overnight admissions since the urgency of admission carries different risks on different days. An analysis including only emergency admissions, which several respondents requested, provides answers which are remarkably similar to the overall result (including some 81% of the deaths) although it excludes some very ill patients (Figure). However the supportive analysis of the subgroup of emergency patients clearly demonstrates that this decision did not drive the results and is remarkably similar to the analysis for the emergency group published in 2012 [1]. This finding will not be surprising to those with some knowledge of epidemiology.
Several correspondents have requested that we share ‘the raw data’, and while we would welcome the possibility of doing this, for governance reasons it is simply not possible. However, these data are available from the HSCIC to those with adequate provision for data management and security. Large datasets such as the one used for our analyses (≈150m) observations require specialist computer equipment and software to setup and undertake the analyses.
It would have been tempting simply to let the trickle of responses run their course, but for the comments that they formed a kind of ‘informal peer review’, and that taken collectively they undermine the validity of our analyses. In a quite remarkable editorial in JAMA, Drummond Rennie observed that the stream of correspondence complaining about the validity a published article contained only letters, and simply receiving letters did not constitute grounds for rejecting important clinical findings.[2] None of the comments raised in the rapid responses undermine the validity of our findings. It is curious to observe correspondents making light of a large number of unexplained deaths over the weekend period instead of these fuelling a consideration on the optimisation of service provision to meet patient need.
References
1. Freemantle N, Richardson M, Wood J, Ray D, Khosla S, Shahian D, Roche WR, Stephens I, Keogh B, Pagano D. Weekend hospitalization and additional risk of death: An analysis of inpatient data. Journal of the Royal Society of Medicine, 2012; 105: 74-84.
2. Rennie D. Smoke and Letters. JAMA. 1993;270(14):1742-1743.
Competing interests: No competing interests