Review of acute pain services

Clinical bottom line

The thrust of this review of the evidence on acute pain services is that they probably improve pain, may reduce common unwanted effects, but may result in more cases of rare but serious harm. Acute pain services probably cost more. The evidence for acute pain services might be described as mixed.

Reference

MU Werner et al. Does an acute pain service improve postoperative outcome? Anesthesia & Analgesia 2002 95: 1361-1372.

Systematic review

Papers published up to February 2001 were sought from reference lists, and textbooks were searched for audits, surveys and clinical trials relating to the use of acute pain services in hospitals.

Results

There were 44 audits and four clinical trials reporting on 84,000 patients. Quantitative synthesis of data was not attempted, and results were reported descriptively. Studies often had a before-after design.

Pain

Most studies reported less pain at rest (by 0-27%) and on movement (by 19-64%).

Nausea and vomiting

Postoperative nausea and vomiting was less frequent in some, but not all studies.

Sedation

There was some evidence that postoperative sedation could be less at hospitals with an acute pain service.

Adverse events

The evidence on adverse events was mixed and not easy to précis. There was a tendency for some to be lower when an acute pain service was in operation. Others may have increased because of the use of different analgesic techniques, like epidurals. Specific cases of serious harm, with denominators, are shown in Table 1.

Table 1: Cases of serious harm recorded

Event description	Number of cases	Total number	Risk
Cauda equina with epidural	1	5602	1:5602
Meningitis with epidural	2	2287	1:1144
Intr avascular migration	3	1062	1:354
Intradural migration	5	4958	1:992
Potential severe complications of infusion device	16	3016	1:189
Accidental epidural opioid overdosing	2	2827	1:1414
Accidental PCA overdosing	3	2922	1:974

Cost issues

There was some evidence that an acute pain service might reduce the length of stay, though other issues might be more important in determining length of stay. The evidence on cost-effectiveness and benefit is mixed. Studies suggest costs per patient ranging from cost savings to about $240.

Comment

The authors have done a great service by bringing all these studies together, telling us what studies have been done, and trying to make sense of the information contained in them. This is a much harder job than for clinical trials with similar patients having similar interventions and where similar outcomes are measured.

The bottom line is that the evidence we have is mixed, and that much more work needs to be done. The authors suggest a focus on the integration of an acute pain service and multimodal rehabilitation techniques, and how that affects outcome in specific procedures. Optimised services may then be tested in large randomised trials, if that were appropriate (though other study architectures may also present themselves).