Wednesday, October 15, 2008

Duration of hospital stay following birth in England

The cesarean news to come out of this year's NHS Maternity Statistics had nothing to do with England's latest national rate (it rose only slightly to 24.3% from 24.1% in 2005-06). Rather, it was the fact that "recovery time in hospital following caesarean section is lessening." The first thing to say is that this statement refers to ALL cesarean deliveries - both elective and emergency. So as usual, it is necessary to go beyond the headlines and seek the truth about the hospital stay of planned surgical procedures - in particular, to discover how this compares to planned vaginal deliveries.

What the NHS summary says
"Approximately one quarter (27,407) spent four days or more recovering in hospital in 2006-07, compared to just under a third (31,393) in 2005-06."

I found this interesting on two counts. Firstly, from the perspective of cost (the less time women spend in hospital, the lower the delivery bill), and secondly, from the perspective of postnatal care (the number of days women spend in hospital following a vaginal delivery is far fewer than that of the previous generation of mothers, but is that necessarily a good thing?).

Short versus long hospital stay
Understandably perhaps, many women dislike spending any more time than is absolutely necessary in hospital. Noisy wards, the risk of infection, hospital food and craving home comforts are just a few of the reasons often cited. Yet for some, especially those women who have just delivered their first baby, they are grateful for the opportunity to have nurses, doctors and midwives on hand to offer advice about feeding and caring for their baby. Or they want the chance to have a few hours rest while their baby is looked after in the nursery. If the luxury of a private room is available, this is an additional bonus.

The implications of cost to the NHS
In many traditional models of cost, where attempts are made to compare the cost of vaginal delivery with cesarean delivery, the duration of hospital stay for a spontaneous vaginal delivery is compared with that of all cesarean deliveries. Even when elective cesarean deliveries are separated out, researchers do not always factor in the duration of stay of all other vaginal delivery outcomes (such as instrumental or emergency cesarean). Clearly, this distorts any financial comparison that seeks to inform birth 'plans' (where the birth outcome is still unknown).

Current cost analysis is flawed
Particularly when it comes to evaluating the cost of maternal request cesarean deliveries. In NICE's 2004 Clinical Guideline, the authors admit that the "estimated cost of maternal request can change depending on the cost value entered in the model.' For example, if the lowest vaginal birth costs reported in the review and the highest CS cost estimate reported in the review are used, the additional cost for accepting 8,747 maternal requests for CS is around £21.2 million.' The report continues: ' Since the highest cost for vaginal birth in the review is higher than the lowest cost for CS, if these values were entered into the model, the model would show that increasing planned CS due to maternal request would lead to savings, which is not a realistic conclusion.' This highlights the fundamental problem in so much of existing birth comparison literature: it works backwards from a preexisting viewpoint or conclusion, and focuses on making the data fit in with it rather than being open to the data leading us to a new reality.

How long do women stay in hospital following different births?
Let's look at some of the figures contained in Table 21: "Duration of delivery episode by method of onset of labour and method of delivery". The majority of women who have a spontaneous vaginal delivery spend 1 or 2 days in hospital (67.6%). 17.6% go home on the day they deliver, and another 14.7% stay for 3 days or more. The majority of women who have a planned cesarean spend 2 or 3 days in hospital (61.2%). 2.9% go home sooner than that, and the remaining 35.9% stay for 4 days or more. But as we know, this is not the comparison that needs to be made. We also need to look at how many days women who 'planned' a spontaneous vaginal delivery, but had a different delivery 'outcome', spent in hospital.

296,058. This is the number of women whose labor began either spontaneously or was induced, and whose delivery is listed as 'spontaneous'.
53,703. This is the number of women whose labor began either spontaneously or was induced, and whose delivery is listed as 'instrumental'.
57,039. This is the number of women whose labor began either spontaneously or was induced, and whose delivery is listed as 'cesarean'.
406,800. This is the number of women who might be described in Table 21 as having 'planned' a vaginal delivery, which means 13.2% needed instrumental assistance and 14% needed an emergency cesarean delivery.

As expected, women with instrumental vaginal and emergency cesarean deliveries had longer hospital stays than women with spontaneous vaginal deliveries. With instrumental deliveries following spontaneous onset of labor, hospital stay is still mainly 1 or 2 days ( 64.7%), but 32.4% stay 3 or more days. Following labor induction, hospital stay for instrumental deliveries is mainly 2 or 3 days (51.6%), but as many as 36.7% of women stay for 4 or more days.

Similarly, with emergency cesarean delivery. Following spontaneous onset of labor, hospital stay is mainly 3 or 4 days (59.4%), with 21.2% staying 5 or more days. And following labor induction, the majority of women spend 3 or 4 days in hospital ( 48.9%), but as many as 44% stay for 5 or more days.

Future research
Only when these longer hospital stays following a planned vaginal delivery are taken into account in cost evaluations, and further - when the elective cesarean group is broken down into women who had a planned cesarean for medical reasons (i.e. the medical reason rather than the cesarean surgery itself may require longer hospital stay) and women who personally decided to have a planned cesarean - will we truly be able to discuss cost implications of cesarean delivery on maternal request.
Additional cost factors for the NHS such as long-term pelvic floor repair or litigation following injuries or death during planned vaginal deliveries is a topic I'll leave for another day...

No comments: