Effect of multifaceted intervention promoting early switch from intravenous to oral acetaminophen for postoperative pain: controlled, prospective, before and after study

Outline of problem

Propacetamol (Pro-Dafalgan, UPSA Laboratories, Rueil-Malmaison, France) is a soluble acetaminophen prodrug for intermittent intravenous infusion (2 g of propacetamol release 1 g of acetaminophen). This drug is often used initially to treat postoperative pain in France, before a switch to oral analgesic drugs.

Propacetamol and paracetamol have identical kinetics and adverse effects, except for possible local pain at the injection site.8 However, administering propacetamol could entail a risk of allergic reactions such as contact dermatitis for health staff.9 Guidelines on postoperative pain management developed by our institution, Assistance Publique-Hôpitaux de Paris, recommend switching from propacetamol to paracetamol as soon as possible.10 In addition, clinical guidelines on acute pain management in adults published in the United States recommend oral administration of drugs as soon as patients can take them.11

Hospital management and medical community representatives strongly recommended cost effective drug use. In view of the high percentage of inappropriate propacetamol administrations, an expert panel was convened (an anaesthetist of the orthopaedic department, two pharmacists, and a public health doctor) in order to improve pain management practices of the orthopaedic department.

Outline of context

This study was performed in Cochin Hospital, a teaching hospital with 1000 acute beds of the Assistance Publique-Hôpitaux de Paris group (the public hospital network for the Paris metropolitan area). The orthopaedic department of this hospital includes 37 beds, 11 surgeons, five consultant anaesthetists (responsible for postoperative pain management), and 30 nurses. In this department 1200 patients undergo surgery each year.

The Therapeutic Committee of the hospital agreed to the study. According to French policy, since the intervention was aimed only at doctors and nurses, it was not necessary to seek patient consent.

Details of approach taken

The general approach was to develop, implement, and assess the impact on practice of a local guideline focusing on both prescription and administration of drugs. Since the guideline would be used as a cost control strategy, we considered it important that an experimental or quasi-experimental design be used to verify that the guidelines did improve care.

We conducted a prospective, controlled, before and after study over 16 months. A pre-intervention observation period (September 1997 to March 1998) was followed by an intervention period (April 1998 to December 1998). The orthopaedic department served as the intervention group, and other surgical departments (general and digestive surgery, urology) of the hospital were the control group. Anaesthetists and nurses of control departments were unaware of the intervention and did not participate in any activity of the orthopaedic department.

Guideline development

In the orthopaedic department, the standard procedure for pain management was to start propacetamol immediately after surgery. Propacetamol injections were maintained as long as the patient felt pain. The guideline simply proposed to switch from propacetamol to oral analgesic as soon as the patient was able to eat solid food or take other oral drugs.

Paracetamol-dextropropoxyphene was preferred to paracetamol alone at the time of switch because it was the most prescribed oral drug for postoperative pain in the orthopaedic department. Too many changes could result in less compliance with the guideline. This combination (a level 2 analgesic according to the World Health Organization's classification12) is considered at least as effective as paracetamol alone (a level 1 analgesic)13 and is recommended by Assistance Publique-Hôpitaux de Paris guidelines when paracetamol alone is ineffective.10

The guideline also included comparison of daily costs of propacetamol (Fr80) and oral drugs (Fr1.30).

Intervention

We implemented the local guideline in various ways.

Fig 1

Poster presenting guideline for early switch from intravenous to oral acetaminophen for management of postoperative pain

• Download figure
• Open in new tab
• Download powerpoint

• For doctors, we used a process of local consensus. Prescribing anaesthetists approved the new guideline during one of their daily staff meetings. The guideline was also formally approved by the head of the orthopaedic department. To allow nurses to switch from intravenous to oral analgesic, the anaesthetist responsible for a patient had to write on the patient's chart, each time propacetamol was ordered, “switch to oral analgesia as soon as the patient eats or takes other oral drugs.”

• For nurses, a one hour educational presentation was made by the expert panel. Four meetings were held in April 1998; to allow all nurses to attend one of them, the meetings were held at different times of day. During these meetings, the expert panel discussed the guideline, compared the effectiveness and cost of paracetamol-dextropropoxyphene and propacetamol infusions, showed the potential cost savings for the department, and answered all questions about the guideline and potential barriers to its application. Nurses' lack of experience in managing narcotic analgesics and lack of confidence in their ability to manage pain play an important role in their behaviour.14 For example, the notion that intravenous medication is better than oral medication is common among patients and health professionals. Thus, the expert panel pointed out that the slow release of acetaminophen from the prodrug was responsible for the similar kinetics of the two routes and that, according to Assistance Publique-Hôpitaux de Paris guidelines, the combination of paracetamol and dextropropoxyphene is more effective than paracetamol alone.

• Posters presenting the new guideline and signed by the management of the department (the surgeon, head of the orthopaedic department, chief nurse, and chief anaesthetist) were displayed in all the nurses' offices after the meetings (figure 1).

• Feedback on practice was provided six months after the guideline was implemented. The expert panel presented preliminary results to nurses and anaesthetists in four meetings and answered all questions raised by the participants.

Measurement of problem

We counted the number of propacetamol injections per patient and calculated the percentage of inappropriate injections during four audits performed in the orthopaedic department at one month before and at one, three, and six months after the intervention was implemented. During each audit (which covered three weeks of clinical practice) one investigator (CR) analysed prospectively medical and nursing records of all patients from 12 hours after their surgical procedure until intravenous analgesic was switched to oral analgesic. When the investigator could not determine the appropriateness of the switch from reviewing the records, she questioned the patients. An inappropriate injection was defined as a propacetamol injection given when a patient was receiving any kind of solid food or oral drug. All data were rendered anonymous and handled confidentially.

In addition, we recorded the monthly consumption of propacetamol injections per 100 hospitalised patients from eight months before to eight months after the intervention was implemented, both in the orthopaedic department (intervention) and in the other surgical departments (control). These data were derived from the pharmacy computerised database and from the hospital information system.

We calculated the standard direct costs of one injection of propacetamol and of one administration of paracetamol, including the costs of the drug and of the medical device used for the injection, and time spent by the nurse (estimated at four minutes for propacetamol and one minute for paracetamol). These costs were drawn from the hospital's financial database. To estimate savings attributable to implementing the guideline, we applied these costs to the average number of injections per patient observed during the audit immediately before and after the intervention. Assuming that the practices observed during these two audits were representative of the practice in the orthopaedic department, we estimated the annual cost savings that could result from the application of the new guideline. We also estimated the cost of the intervention, including labour cost.

Statistical analysis—We used a two tailed formulation to test the null hypothesis that the monthly use of propacetamol injections per 100 patients was unchanged by the intervention. We performed Student's t tests to compare use of injections before and after the intervention and calculated 95% confidence intervals.