Abstract

Background

To evaluate the association between daily morphine requirement and the intravenous patient-controlled analgesia (IVPCA) setting in patients undergoing spinal surgery.

Methods

We conducted a retrospective analysis of 179 patients of American Society of Anesthesiologists physical status class I-III who underwent elective posterior lumbar spinal surgery and consented to IVPCA for postoperative pain control. The regi-mental solution contained morphine 1 mg/mL. The IVPCA program was set to deliver a priming dose of 1.5–4 mL, a basal infusion rate of 0–1.2 mL/hr, and a 0.5–1.5 mL bolus on demand with a 5-minute lockout interval. Demographic data, surgical procedures, analgesia program setting variables, 4-hour cumulative morphine dose and 11-point numeric rating scale for pain on postoperative days 1 and 2 were collected for comparison.

Results

The IVPCA requirement decreased gradually over time (p < 0.001). The number of vertebrae involved significantly influenced the daily morphine requirements (p = 0.01). None of the IVPCA settings, including continuous infusion, affected daily morphine requirements. On average, the analgesic requirement on postoperative day 2 was 18% less than that on postoperative day 1.

Conclusion

The number of vertebrae involved was significantly associated with the daily IVPCA requirement. The IVPCA settings, including priming dose, basal infusion rate and bolus dose, did not affect the daily morphine requirements.

Keywords

general anesthesia; intravenous patient-controlled analgesia; morphine; spinal surgery;

1. Introduction

It is well known that pain after spine surgery is severe and tends to last for at least 3 days postoperatively. ¹⁻³ Opioids are excellent drugs for pain control and morphine is the most extensively used agent, administered by intramuscular injection, intravenous injection, or epidural application. Intravenous patient-controlled analgesia (IVPCA) was introduced in the late 20th century⁴⁻⁶ and has gradually gained popularity over conventional intramuscular analgesia given at fixed intervals. PCA is now widely used for patients undergoing spinal surgery, ranging from simple decompression and vertebroplasty, to bone graft fusion, instrumentation, and correction of scoliosis.^1−3,7 However, there are no recommendations on how to set the IVPCA variables for spinal surgery. Although the current IVPCA settings in our institution were derived from clinical experience gained in other surgical procedures with similar pain scores, we hypothesized that the IVPCA pump settings are not associated with the patients’VPCA demand. Therefore, we conducted a retrospective survey to evaluate the IVPCA variables such as the use of continuous infusion and the size of the bolus dose on daily morphine requirements for 48 hours postoperatively. Other potential influential factors, including patient and surgical characteristics, were also included in this analysis.

2. Methods

This retrospective study was conducted at Taipei Veterans General Hospital with ethics approval from the institutional review board. The inclusion criteria were patients who underwent elective posterior lumbar spinal surgeries and consented to receive IVPCA for postoperative pain control between March 2005 and January 2006. Exclusion criteria were patients who used IVPCA for < 48 hours for any reason, and those who received postoperative mechanical ventilator support or intensive care.

  The primary endpoint was IVPCA morphine dose on postoperative days 1 and 2. For all patients, general anesthesia was induced with fentanyl (3−5 μg/ kg), thiopental (4−6 mg/kg) and rocuronium (0.8 mg/ kg). Isoflurane or desflurane in 40−60% oxygen were used to maintain anesthesia. At the end of the procedure, the muscle relaxant was reversed and the endotracheal tube was removed upon full recovery of respiratory strength.

  On arrival at the postanesthetic care unit, the IVPCA infusion device (Aim Plus system; Abbott Laboratories, North Chicago, IL, USA) was connected to the patient. The regimental solution contained 1 mg/mL morphine. The PCA program was set to deliver 0.5−1.5 mL infusate on demand with a 5-minute lockout interval and a 4-hour dose limit. The basal infusion rate was set at 0−1.2 mL/hr with a priming dose of 1.5−4 mL as prescribed by the anesthesiologist. All patients were visited at least once daily by clinical staff in the morning or afternoon, and whenever the circumstances required, to check on the patient’s IVPCA status. The analgesic effect of IVPCA was assessed using an 11-point numeric rating
scale (0 = no pain and 10 = worst pain imaginable). The 4-hour cumulative morphine dose on postoperative days 1 and 2 was also recorded. Data collected on postoperative day 3 were excluded because of inconsistent case endpoints. Other data collected included the patients’ sex, age, weight, height, body mass index (BMI), surgical procedure and location.

2.1. Statistical analysis

Parametric data are expressed as mean and standard deviation. Categorical data are presented as n and percentage. Linear mixed models with Toeplitz’s covariance structure were used to evaluate withinsubject time effects and other between-subject effects of repeated measures of IVPCA requirements. ^8,9 Interaction effects were also examined. A p value of < 0.05 was considered statistically significant. Linear regression analysis was conducted to explore the relationships between IVPCA requirements on each day. According to Peat and Barton,¹⁰ the minimum number of cases for multiple regression analysis should be at least 100. Because the sample size was 179 in this study, this criterion was met. All statistical analyses were conducted using SPSS version 15.0 (SPSS Inc., Chicago, IL, USA).

3. Results

A total of 179 patients were included in the analysis. Patients’attributes and data related to IVPCA usage are presented in Table 1. Overall, 57.5% of patients received spinal surgeries involving more than two segments and 84.9% patients underwent instrumentation. The total IVPCA requirements on postoperative days 1 and 2 were 44.2 mL and 36.2 mL, respectively. The cumulative 4-hour morphine doses are shown in Figure 1.

  Table 2 shows the effect of various factors on the daily IVPCA requirement. The IVPCA requirements decreased gradually over time (p < 0.001). The use of instrumentation and the number of spine segments being operated on had significant effects on daily IVPCA requirement. However, the effect of instrumentation disappeared after adjusting for the number of vertebrae involved (p = 0.1). Interactions with time were not significant for any of the factors.

  Table 3 shows the estimates of significant between-subject effects and within-subject effects of time on morphine consumption. Patients receiving decompression for more than two vertebrae tended to consume more morphine than those who underwent decompression of two vertebrae. The association between the number of vertebrae involved and daily IVPCA requirement is illustrated in Figure 2. Moreover, if IVPCA consumption at 44−48 hours was used as a baseline for comparison, we found that the 4-hour-interval IVPCA requirements on postoperative day 1 were significantly higher than that on postoperative day 2. This phenomenon disappeared after 28 hours.

  Figure 3 presents a scatter plot of the IVPCA requirements on postoperative days 1 and 2. The correlation coefficient between the IVPCA requirements on these days was 0.71 (p < 0.001). The average IVPCA requirement on postoperative day 2 was 18% less than that on day 1.

Table 1 Patient and surgical characteristics, and intravenous patient-controlled analgesia usage*

Download full-size image

Download full-size image

Figure 1 Cumulative morphine requirements in consecutive 4-hour intervals. The error bars represent the 95% confidence intervals. IVPCA = intravenous patientcontrolled analgesia.

Table 2 Between-subject effects and interaction with time between patient characteristics and intravenous patient-controlled analgesia requirements on postoperative days 1 and 2

Download full-size image

Table 3 Estimates for the significance of betweensubject and within-subject time effects

Download full-size image

Download full-size image

Figure 2 The relationship between intravenous patientcontrolled analgesia (IVPCA) requirements for each postoperative day and number of vertebrae involved. The error bars represent 95% confidence intervals.

Download full-size image

Figure 3 Scatter plot with linear regression lines for the intravenous patient-controlled analgesia (IVPCA) requirements on postoperative days 1 and 2. The regression line fits without an intercept. This analysis revealed a linear trend, indicating that the IVPCA demand on postoperative day 1 is positive correlated with the IVPCA demand on postoperative day 2. The correlation coefficients for the IVPCA demand was 0.71 (p < 0.001), which means that the IVPCA demand on postoperative day 1 accounts for about 51% (0.712) of the total variability of the IVPCA requirement on postoperative day 2.

4. Discussion

PCA offers patients an opportunity to take part in their pain management through miscellaneous infusion designs. The PCA pump settings include initial loading dose, bolus analgesic dose, lockout interval for bolus doses, basal infusion rate, and maximal dose limit within a specific time period. We found that loading dose, bolus dose and basal infusion rate did not influence the daily IVPCA requirements. It is reasonable to assume that a single priming dose and different sizes of bolus dose did not significantly affect the daily morphine requirement because the optimal analgesia for each individual was achieved by titration. One of the merits of PCA is that it provides a stable and sufficient analgesic level based on the individual’s requirement. Once the optimalmorphine therapeutic range for analgesia has been reached, the demand is no longer related to the priming or bolus doses. On the other hand, previous studies found that the basal infusion rate might increase daily morphine consumption and increase the incidence of adverse effects^.4,11−14 However, we did not find a similar effect of continuous infusion on the daily total morphine dose. One possible explanation for this finding is that the infusion rate used in our study was relatively low compared with that used in previous studies. For example, Owen et al4 used an infusion rate of 1.5 mg/hr, while McCoy et al¹² and Parker et al¹⁴ used a rate of up to 2 mg/hr. The lower PCA infusion rates in our study possibly resulted in weaker analgesic effects and additional patient-determined bolus doses were needed to reach the optimal therapeutic range. Therefore, the low basal infusion rate used in this study had little effect on daily IVPCA morphine consumption. High basal infusion rates are not recommended and should be avoided to prevent adverse effects and loss of the safety benefits of PCA.^15,16

  We also examined the effects on patient characteristics and surgical extent on IVPCA requirements. To eliminate potential confounding factors, we selected patients undergoing similar spine operations to enhance the homogeneity of the study population. The effects of demographic and surgical factors were also evaluated and adjusted if necessary. We found that the daily IVPCA requirement for patients undergoing spinal surgery was related to the number of vertebrae involved in the operation, but not to patient characteristics such as age, height, weight, BMI and sex. These results are consistent with those of other studies, which revealed that weight and height were not associated with morphine consumption.^4,6,14,17 Although a decrease in morphine requirement with increasing age is evident in many studies,^6,14,18,19 we did not find a similar relationship. Indeed, other studies had similar results.^4,17 For example, Chia et al¹⁷ did not find an inverse association between age and morphine consumption in Chinese individuals. Differences in results between these studies might be due to the homogeneity of the study population or that most of our subjects were not old enough to alter their pharmacokinetics of morphine.

  Table 2 lists the potential factors that were considered likely to affect the IVPCA requirements, and their interactions with time. We found that none of these factors had an interaction with time. This implies that their role in IVPCA requirement did not change over time. It is reasonable to hypothesize that the IVPCA requirement decreases gradually with time, as found in other studies.^14,20 However, the pattern of this trend is still unknown. Repeated measures analysis with shorter time intervals may better reveal the effect of time. We analyzed the cumulative dose in 4-hour intervals to avoid random fluctuations associated with shorter intervals. Most of the earlier largescale studies focused on the total dose achieved with IVPCA and did not emphasize the effects of time.^{4,6,11,14,20−22} Our study demonstrated that these between-subject effects remained insignificant throughout the duration of IVPCA. Our data also revealed that the IVPCA morphine requirement decreased over time, with a reduction of 18% between postoperative days 1 and 2.

  There are some limitations in our study. First, the data were collected in a retrospective manner and the potential confounding factors could not be equally controlled. Second, the incidence of morphine side effects was not evaluated.

  In conclusion, despite using various priming doses, bolus doses, and basal infusion rates for PCA, the only factor associated with IVPCA requirement was the number of vertebrae involved. In contrast, age, height, weight, BMI and sex were not associated with IVPCA requirement. The amount of morphine delivered decreased linearly over time and was approximately 18% lower on postoperative day 2 than on day 1.

Acknowledgments

This study was supported in part by grants from Taipei Veterans General Hospital (V97B1-013) and the National Science Council (NSC 96-2314-B-075- 059), Taiwan, R.O.C.