Patient-controlled analgesia (PCA) is a useful method for managing post-operative pain in surgical patients , but controversy still exists about the effects of PCA on morbidity, and cost savings in nursing personnel time and duration of hospitalization [2-4]. It is generally accepted that PCA offers clinical benefits over conventional intramuscular (i.m.) analgesia regimens [5,6], but because of its relative ease of administration, the i.m. route of administration of opioids is still routinely used despite variable results with regard to time of onset, intensity and duration of analgesia .
The cost of treatment has become the subject of a great deal of discussion at many institutions, and has forced new technologies to be evaluated from the perspective of cost-effectiveness and cost-benefit before their incorporation into clinical practice. The purchase of dedicated electronic PCA pump systems represents a significant financial investment. More recently, a disposable non-electronic PCA system has been introduced. Although proposed initially for chronic pain relief for ambulatory patients, one study suggested that, for a variety of reasons, the device was best used by both patients and nurses for acute post-operative pain treatment, in spite of being the most expensive option available . Therefore, it appeared logical to assess such a disposable PCA system from the point of view of cost-benefit and cost-effectiveness as well as for therapeutic efficacy.
The present study contrasted parenteral PCA with a disposable device for i.m. post-operative pain relief in terms of drug, equipment and personnel costs, the demand on nursing time, the analgesic efficacy, the functional recovery and ambulation of the patients, and the duration of their hospital stay.
Subjects and methods
The present study included 40 female patients (ASA physical class I-II), aged 35-69 years, scheduled for abdominal subtotal hysterectomy. All patients gave written informed consent to the study, which was approved by the University Committee on Human Research. Using a computer-generated randomization table, patients were allocated to receive post-operative analgesia by either PCA or i.m. injection. The PCA method used a disposable PCA Infusor System (Basal/Bolus Infusor 15, model F2C1955, Baxter Healthcare Corporation, Deerfield, IL, USA). This consisted of two components: a 65-mL elastomeric balloon reservoir delivering a fixed-rate basal intravenous infusion (0.5 mL h−1) by microbore tubing; and an on-demand bolus device through a second microbore tubing (0.5 mL, 15-min dosing interval) connected to a patientcontrol module worn around the wrist. This device was linked to the intravenous (i.v.) line via a nonreturn valve [9-11].
All patients were instructed in the method of the study and had practice with the use of the visual analogue scale (VAS). They were premedicated with midazolam 0.1 mg kg−1 and glycopyrrolate 0.2 mg i.m. Anaesthesia was induced with propofol 1.5-2.0 mg kg−1 and vecuronium 0.1 mg kg−1, and maintained with isoflurane in a nitrous oxide-oxygen mixture and incremental vecuronium supplements for muscle relaxation. On arrival at the post-anaesthesia care unit (PACU), post-operative pain treatment was started immediately after an initial evaluation of pain intensity. The PCA patients received a continuous i.v. infusion of 1.5 mg h−1 piritramide with incremental doses of 1.5 mg with a lock-out interval of 15 min. The i.m. group received 0.3 mg kg−1 piritramide intramuscularly, followed by 0.3 mg kg−1 increments administered at the request of the patient at a minimum interval of 5 h, i.e. the standard practice for pain treatment on the ward. Time between request and dose administration was less than 10 min. Pain intensity was measured using a VAS  every 15 min for the first hour, hourly for the next 8 h and every 4 h until 72 h post-operatively. Pain intensity difference, i.e. the difference between the actual pain and the pain intensity at the time when the drug was first administered, was multiplied by the fraction of an hour since the previous evaluation and was totalled for the entire observation period to provide a further estimate of analgesia, the sum of pain intensity differences (SPID). The level of sedation was evaluated according to a scale (Table 1). Sedation scores were evaluated over the first 3 post-operative days. The dose of piritramide given daily and the cumulative dose after 3 days was calculated.
Functional recovery was estimated by the gynaecologist and nurses of the ward. The recovery assessments included time to drink and eat, the times to first flatus and stool evacuation, to getting out of bed, to walk in the hospital corridor, and to discharge from the hospital. The nursing staff of the PACU and the ward were instructed to time their nursing activities, which were subdivided in two categories: (1) the time related to post-operative pain treatment; and (2) the time needed for aiding the patient (collection of vital signs, help in positioning and ambulation) and performing hygienic care tasks. A specially designed nursing log was used to note the nursing times, which were measured with a stopwatch. The nursing logs were collected and the times were calculated by the author.
The cost of each therapy was calculated per patient based on cost per ampoule of piritramide, the infusor set and tubing, and ancillary equipment per injection. This included costs of syringes, needles, alcohol swabs and i.v. tubing in both groups. Personnel costs associated with the direct pain management were determined using the actual time (in minutes) a nurse spent evaluating or treating the patient multiplied by the hourly wage paid by the hospital. Nursing labour costs for aiding and general care were also calculated. Hospitalization cost was calculated by multiplying the length of hospital stay by an average per day hospital charge (10 533 BEF day−1).
Total direct costs included the cost of analgesic treatment, and the nursing cost for aiding and general care. From these data, a cost-benefit and cost-effectiveness analysis were performed . The costbenefit analysis measured the cost of analgesic treatment and benefits (nursing time) in monetary units and computed a net monetary gain or loss. The cost-effectiveness analysis compared the two treatment programmes having a common health outcome (analgesia) with costs expressed as monetary units and effectiveness or quality of treatment expressed in arbitrary units of SPID.
Data are presented as the mean ± SEM. The data were analysed by one-way analysis of variance (ANOVA). Statistical significance for intergroup differences was accepted if P<0.05.
Both groups were comparable with regard to demographic aspects (i.e. age, weight and height), duration of surgery and anaesthesia (Table 2).
Quality of analgesia and sedation scores
The initial VASs were similar for both groups. Pain intensity was reduced by 25% after 3 h in both groups, but marked differences were observed after 16 h, and subsequently, when the reduction in pain intensity was consistently better in the PCA group at all observation times (P<0.01) (Fig. 1).
The analgesic efficacy of the PCA infusor system is also reflected in a significantly higher SPID value (PCA group=4109±215, i.m. group=2877±312; P<0.001). Sedation scores were similar in both groups. Sedation diminished progressively during the immediate postoperative period, and no patients in either group showed meaningful sedation after 24 h, despite the larger amounts of piritramide used in the PCA group (Fig. 2).
On the day of surgery, the dose of piritramide used was similar for both groups (Table 3). During the first and second post-operative days, the average amounts of piritramide used were significantly lower in the i.m. group, as was the total dose of narcotic analgesic consumed over the 3-day observation.
Functional recovery of the patients
The times to first drinking, eating, and voiding flatus or stool evacuation were similar for both groups (Table 4). While no difference was observed for the time to sitting in a chair (or leaving the bed), the patients in the i.m. group were able to ambulate earlier in their room and ward.
Influence on nursing activities
The PCA group consistently needed less nursing time for post-operative pain treatment during the 3-day observation period with significant reductions during the day of surgery and the first post-operative day (Table 5). General hygienic care and nursing help required less nursing time in the PCA group during the day of surgery and post-operative day 1 as compared with the i.m. group, but these differences were not significant.
Cost of treatment
Table 6 summarizes the average global cost of treatment expressed in monetary units (Belgian Franc, BEF) for the two groups of patients. The i.m. treatment was less costly in respect of drug and equipment expenses, but carried a markedly higher cost for nursing time (P<0.001). The costs of nursing help and care were lower in the PCA group, but did not decrease the total direct cost when compared with the i.m. group (P<0.001).
The cost-benefit analysis indicates a higher cost-benefit ratio for the PCA treatment (1.1 versus 0.35 for the i.m. treatment). The total analgesic therapy cost represented 59% of the total direct costs in the PCA group and 33% in the i.m. group. The average cost difference in total analgesic therapy amounted to 2904 BEF to the disadvantage of the PCA treatment and was not compensated for by the differences in costs of nursing help/care. The impact of the i.m. treatment on nursing resources for the treatment of pain, help and care for the patients was larger and represented an average direct cost of 4725 BEF compared with 4110 BEF for the PCA treatment. The total direct costs of both treatments were significantly different (P<0.001).
With regard to cost-effectiveness, the PCA treatment provided more effective analgesia than the i.m. treatment. Assuming that the quality of treatment with conventional i.m. therapy was graded 1, the quality of treatment with PCA therapy was 1.43, i.e. 43% higher, based on the SPID. Considering SPIDs as units of non-monetary effectiveness, the cost of analgesic therapy and total nursing costs per unit effectiveness amounted to 1.1 and 1.0 BEF, respectively, for PCA treatment, and 0.6 and 1.6 BEF, respectively, for conventional i.m. therapy. However, the total direct costs per unit effectiveness were similar (PCA group BEF = 1.9 BEF, i.m. group = 1.7 BEF).
This prospective evaluation of PCA and i.m. modes for post-operative pain treatment confirmed the better efficacy of analgesia provided by the non-electronic Basal/Bolus infusor system. The unblinded study design may have resulted in bias in the VASs and nursing times. However, a double-blind, double-dummy design was too difficult to implement for recording utililization of nursing resources and might have had an impact on cost estimation. The present authors felt that strict randomization of a homogeneous group of patients (i.e. all submitted to a standardized surgical procedure) was the fundamental requirement for efficacy and cost evaluation.
The safety and effectiveness of this non-electronic PCA system has been previously documented (10, 14-16). Ballantyne et al. found that patients treated with PCA had an additional benefit of 5.6 points on the 0-100 VAS. In the present study, only slight differences in pain score were found during the first 12-h post-operatively. This may have been because the nurses in the PACU were familiar with treating pain adequately. However, pain relief was significantly better from the sixteenth hour post-operatively in the PCA patients. This observation could be explained by the fact that PCA patients received a background infusion of piritramide in addition to being able to adjust their dosing to their pain perception. Intramuscular injections of analgesics characteristically result in unpredictable plasma concentrations with two-fold variations in drug peak concentrations and three- to seven-fold variations in time to peak concentrations . Moreover, the doses of prescribed i.m. analgesics may be too small for optimal pain relief and nurses tend to cut down further on the quantity or the frequency of intramuscular opioid administration . Individual analgesic requirements were reported to be lower in PCA-treated patients than following conventional i.m. treatment [2,20]. However, recent studies did not confirm this observation [6,21].
After abdominal surgery, the return of bowel function and oral intake are important steps in the patient's recovery. Once ileus has resolved, enteric nutrition can begin and toleration of a regular diet is the prerequisite for most patients to eventual discharge from hospital. The present results showed no significant difference in the recovery of bowel function, although the dose of piritramide administered was higher in the PCA group than in the i.m. group. Whether the technique of pain treatment has a predictable influence on bowel function is debatable. Although Jackson et al. reported a shorter duration of post-operative ileus, delayed resolution of adynamic ileus was observed after radical retropubic prostatectomy  and caesarean section  during PCA therapy.
Other assessments of functional recovery of patients, such as the time to oral intake of fluids and food, and the ability to leave the bed showed no differences between the two analgesic treatments. However, patients treated with conventional i.m. therapy could ambulate sooner than patients receiving PCA therapy. Passchier et al. found that PCA led to more fatigue and less vigour, which might explain our observation.
Ballantyne et al. calculated that the mean difference in the duration of hospitalization stay was 4 h longer with conventional analgesia. In the present study, this amounted to 10 h, but the decisions were largely dependent on surgical indications.
The cost aspects of the two analgesic therapy techniques were examined using two approaches: cost-benefit and cost-effectiveness analysis. Cost-benefit analysis defined both benefits and costs in monetary terms. The objective was to determine if the cost saving realized in nursing time outweighed the costs incurred from the therapy. The hospitalization cost was not included in the analysis since the present authors considered that the analgesic therapy only played a limited role in the duration of hospital stay for the selected type of surgery. The present authors expressed the benefit as the functional recovery, which was almost identical in both groups, and the nursing costs related to help and care of the patients. From this analysis, it was concluded that the more expensive PCA treatment had a higher cost-benefit ratio. The reduced nursing costs associated with the PCA pain treatment did not compensate for the higher costs of equipment and drugs. In a cost-effectiveness analysis, effectiveness is expressed in terms of non-monetary outcomes and the resources used associated with the therapies are compared. In this analysis, both the direct medical costs (e.g. drugs and equipment), and the costs associated with caregivers were considered and balanced against the efficacy of the two treatments. The i.m. treatment had a lesser analgesic efficacy and a larger impact on nursing costs. When considering total direct costs, both therapies resulted in almost identical cost-effectiveness (1.9 BEF for PCA treatment vs. 1.7 BEF for i.m. treatment). However, the total price of pain treatment and direct nursing costs are relatively small in contrast with the calculated hospitalization costs (Table 6).
In conclusion, the PCA and i.m. treatments gave comparable analgesic efficacy during the initial post-operative period, but PCA provided better analgesia from the sixteenth hour onwards, a result which was probably related to the administration of a higher dose of piritramide. Functional recovery was similar in both groups with the exception of the time to ambulation, which was shorter after conventional i.m. analgesic treatment. Nursing time for pain treatment was significantly reduced during PCA treatment. Cost analysis showed that PCA treatment had a higher cost-benefit ratio. However, because of the more efficacious analgesia of PCA, cost-effectiveness was similar for both analgesic treatments.
The authors thank the nursing staff of the PACU and the gynaecological ward for the meticulous timing of their nursing activities
1 Owen H, Mather LE, Rowley K. The development and clinical use of patient controlled analgesia. Anaesth Intensive Care
2 Albert JM, Tallbott TM. Patient controlled analgesia vs. conventional intramuscular analgesia following colon surgery. Dis Colon Rectum
3 Hecker BR, Albert I. Patient-controlled analgesia: a randomised prospective comparison between two commercially available PCA pumps and conventional analgesic therapy for postoperative pain. Pain
4 Ross EL, Perumbeti P. PCA: Is it cost effective when used for postoperative pain management? Anesthesiology
5 Wasylak TJ, Abbott FV, English MJM, Jeans ME. Reduction of postoperative morbidity following patient controlled morphine. Can J Anaesth
6 Boulanger A, Choiniere M, Roy D, Boure B, Chartrand D, Choquette R, Rousseau P. Comparison between patient-controlled analgesia and intramuscular meperidine after thoracotomy. Can J Anaesth
7 Mather LE, Lindop MJ, Tucker GT, Pflug AE. Pethidine revisited: plasma concentration and effects after intramuscular injection. Br J Anaesth
8 Sawaki Y, Parker RK, White PF. Patient and nurse evaluation of patient controlled analgesia delivery systems for postoperative pain management. J Pain Symptom Management
9 Irwin M, Gillespie JA, Morton NS. Evaluation of a disposable patient-controlled analgesia device in children. Br J Anaesth
10 Mackey NA, Ilsley AH, Owen H, Plummer JL. Laboratory evaluation of the Baxter patient controlled analgesia infusor system: a disposable patient controlled analgesia device. Anesth Analg
11 Wermeling DP, Foster TS, Rapp RP, Kenady DE. Evaluation of a disposable, non-electronic, patient-controlled analgesia device for postoperative pain. Clin Pharm
12 Revill SI, Robinson JO, Rosen M, Hogg MIJ. The reliability of a linear analogue for evaluating pain. Anaesthesia
13 Vitez TS. Principles of cost analysis. J Clin Anesth
14 Robinson SL, Rowbotham DJ, Mushambi M. Electronic and disposable patient controlled analgesia systems. A comparison of the Graseby and Baxter systems after major gynaecological surgery. Anaesthesia
15 Davidson JAH, Dryden CM, Smith DD. Laboratory assessment of the Baxter disposable patient controlled analgesia system. Anaesthesia
16 Rowbotham DJ, Wyld R, Nimmo WS. A disposable device for patient-controlled analgesia with fentanyl. Anaesthesia
17 Ballantyne JC, Carr DB, Chalmers TH, Dear KBJ, Angelico IF, Mosteller F. Postoperative patient-controlled analgesia: meta-analyses of initial randomised control trials. J Clin Anesth
18 Austin KL, Stapleton JV, Mather LE. Multiple intramuscular injections: a major source in variability in analgesic response to meperidine. Pain
19 Marks RM, Sacher EJ. Undertreatment of medical inpatients with narcotic analgesia. Ann Intern Med
20 White PF. Use of patient controlled analgesia for management of acute pain. JAMA
21 Dahl JB, Daugaard JJ, Larsen HV, Mouridsen P, Nielsen TH, Kristoffersen E. Patient controlled analgesia: a controlled trial. Acta Anaesthesiol Scand
22 Jackson D. A study of pain management: patient controlled analgesia vs. intramuscular analgesia. J Intravenous Nursing
23 Stanly BK, Noble MJ, Gilliland C, Weigel JW, Mebust WK, Austenfeld MS. Comparison of patient-controlled analgesia vs. intramuscular narcotics in resolution of postoperative ileus after radical retropubic prostatectomy. J Urol
24 La Rosa JA, Saywell RM, Zollinger TW, Oser TL, Erner BK, McClain E. The incidence of adynamic ileus in post-cesarean patients. Patient-controlled analgesia vs. intramuscular analgesia. J Reprod Med
25 Passchier J, Rupreht J, Koenders MEF, Olree M, Luitwieler RL, Bonke B. Patient-controlled analgesia leads to more postoperative pain relief, but also to more fatigue and less vigour. Acta Anaesthesiol Scand