Monday, May 8, 2006
Archives of Internal Medicine has just published a large trial of a project to improve pain management in the hospital from R. Sean Morrison & Diane Meier & colleagues at Mt. Sinai in New York. It was a somewhat complicated trial involving different combinations of pain interventions amongst hospitalized medical, surgical, and oncology patients at Mt. Sinai. Click on the table below to see the interventions in the two blocks (block A was one set of wards and block B was another--all patients admitted to these wards received the interventions, although data was collected only on those who consented etc.).
As you can see the 3 main interventions were 1) use of an enhanced pain scale (current severity, worst pain, pain relief, pain acceptibility vs assessing current severity only on a 0-3 scale), 2) nursing audit and feedback about pain assessment, and 3) a computerized clinical decision support tool (the exact details of which remain murky--it appears to be something developed just for this trial and integrated into the Mt. Sinai computer system--it offered suggestions and prompts about appropriate analgesic prescribing but was a passive/ignorable system).
Assessment-wise, the use of the enhanced scale increased the amount of nursing pain assessments, and the use of audit-feedback or the computer system increased nursing assessment even more. To give you a sense of the magnitude--rates of pain assessment went from ~20-30% to ~60-70% with the interventions. The interventions also increased the percent of patients with moderate to severe pain who received a WHO Step 2-3 analgesic (i.e., an opioid), although pain severity was not affected by any of the interventions. The mean pain level of the patients on days 1-3 was however 1/3 (=mild) indicating that this may not be the best group of people to test pain interventions on.
Studies like this are also nice opportunities to look at the natural history of pain in the hospitalized patient--in this case 30% of patients had moderate to severe pain at enrollment--which I would bet makes it one of the most common symptoms experienced in the hospitalized population.
Anyway, what to make of this? I'm confused by the exact nature of the computerized clinical decision support tool--how that improved nursing assessment of pain on par with auditing their performance--but that may just be due to the way it is described in this article. Regardless it, along with auditing, quite dramatically improved nursing assessment of pain--that is good news and a challenge to any institution to do the same. The big disappointment of the trial is I guess the not-too-surprising lesson that physician behavior is tough to change. Maybe that is too harsh--this intervention did change physician behavior--it increased prescription rates of analgesics. Perhaps they just didn't do a good job of analgesic titration? One hopes the Mt. Sinai group continues this line of investigation and develops an intervention that demonstrably improves pain control and physician pain management in the hospital. (And one then also hopes it would be easily portable to other hospitals). An obvious solution would be using audit-feedback on the physicians, but one can imagine a certain amount of 'resistance' to that if you know what I mean.