|Home | About | Journals | Submit | Contact Us | Français|
We previously reported that continuous perineural femoral analgesia reduces pain with movement during the first 2 days after anterior cruciate ligament reconstruction (ACLR, n=270), when compared with multimodal analgesia and placebo perineural femoral infusion. We now report the prospectively collected general health and knee function outcomes in the 7 days to 12 weeks after surgery in these same patients.
At 3 points during 12 weeks after ACLR surgery, patients completed the SF-36 General Health Survey, and the Knee Outcome Survey (KOS). Generalized Estimating Equations were implemented to evaluate the association between patient-reported survey outcomes and (i) preoperative baseline survey scores, (ii) time after surgery, and (iii) 3 nerve block treatment groups.
Two-hundred-seventeen patients’ data were complete for analysis. In univariate and multiple regression Generalized Estimating Equations models, nerve block treatment group was not associated with SF-36 and KOS scores after surgery (all with P≥0.05). The models showed that the physical component summary of the SF-36 (P < 0.0001) and the KOS total score (P < 0.0001) increased (improved) over time after surgery and were also influenced by baseline scores.
After spinal anesthesia and multimodal analgesia for ACLR, the nerve block treatment group did not predict SF-36 or knee function outcomes from 7 days to 12 weeks after surgery. Further research is needed to determine whether these conclusions also apply to a nonstandardized anesthetic, or one that includes general anesthesia and/or high-dose opioid analgesia.
Implications Statement: After anterior cruciate ligament reconstruction (ACLR) with spinal anesthesia and multimodal analgesia, patients were evaluated 3 times from 7 days to 12 weeks after surgery with (i) the SF-36, and (ii) the Knee Outcome Survey–Activities of Daily Living Scale. There were no differences in general health or knee function based on nerve block treatment group; continuous femoral nerve block analgesia was not associated with any adverse effects on patient-reported general health or knee function in the 7-day to 12-week period after ACLR.
Patient outcomes after anterior cruciate ligament reconstruction (ACLR) have not been well described in the first few days and weeks after surgery. Continuous femoral perineural analgesia helps sustain knee analgesia during these few days1 when included with spinal anesthesia and multimodal analgesia. However, general health and physical function outcomes up to 12 weeks after ACLR with this (or any) anesthetic-analgesic technique have not been reported. After total knee replacement 2,3, patients demonstrated better physical function up to 6 weeks after surgery when continuous femoral perineural analgesia was used. Therefore, we evaluated patient outcomes between 7 days and 12 weeks after ACLR, to test the hypothesis that continuous perineural femoral analgesia (versus its absence) leads to the specific aims of improved physical function and general health status, as has been demonstrated after total knee replacement.
After achieving approval by the IRB of the University of Pittsburgh Medical Center and obtaining informed consent, ACLR patients underwent a standardized multimodal anesthesia and analgesia regimen (previously described) that included an ipsilateral hyperbaric spinal with bupivacaine.1 These patients consented for all described study procedures and evaluations both before surgery (e.g., baseline data collection), during the day of surgery (e.g., nerve block treatment group assignment), and in the 12 weeks of follow-up after surgery (patient-reported outcome surveys described below).
Patients were eligible for participation if they were 14 to 65-years-of-age and had an ASA physical status (ASA/PS) of I-II. The patients included in this 7-day to 12-week outcomes analysis (after surgery) underwent outpatient ACLR with a single-bundle allograft, single-bundle patellar tendon autograft or a single-bundle hamstring autograft. Two-hundred-seventy patients were recruited, and exclusions after enrollment were not replaced. Patients were excluded from study enrollment based on variables previously defined in our earlier publication.1
Patients were randomized to one of the following femoral nerve block catheter treatment groups after surgery was performed: (i) saline bolus (30 mL) plus saline infusion (270 mL at 5 mL/hr, group SbSi); (ii) levobupivacaine (0.25%) bolus with saline infusion (group LbSi), or (iii) levobupivacaine (0.25%) bolus plus levobupivacaine (0.25%) infusion (group LbLi). These details, along with the detailed descriptions of the perioperative multimodal analgesic and antiemetic care protocol, are listed in a previous manuscript.1
The SF-36 4,5 is generally accepted as a “gold standard” of assessing general health status. The SF-36 has been used frequently to assess patients with ACL-deficient knees 6–8, and co-administration of the SF-36 with other knee function surveys has been recommended.9 The domains that the SF-36 measures are as follows: physical function, role physical, bodily pain, general health, vitality, social function, role emotional and mental health. Through a standardized norm-scoring process, these 8 dimensions are weighted and transformed into 2 summary scores: the physical component summary (PCS) and the mental component summary (MCS).
The Knee Outcome Survey-Activities of Daily Living Scale (KOS-ADLS) 10 was developed and validated to evaluate patient-reported knee function. The 2 primary dimensions comprising the total score are symptoms (6 separate items, including pain and weakness as 2 items), and functional limitations (8 separate items, including pain during kneeling). All items in these 2 dimensions are graded by respondents on a 6-point Likert scale ranging from “Activity is not difficult” to “I am unable to perform the activity.” A separate 3-item section is included for patients to grade their global knee function assessment in 3 ways: an analog functional score between zero and 100 and 2 items featuring 4-choice Likert scale responses ranging from “severely abnormal” to “normal.”10
During the preoperative interview and after informed consent for the study was obtained, baseline survey data (SF-36 and KOS-ADLS) were obtained, as was additional demographic information (age, gender, ethnicity, height, weight, ASA/PS classification, smoking status). Surgeon, tourniquet use, tourniquet time and case duration were also recorded during the intraoperative course.
At the start of the study, the surgeons’ office visits for the patients were to take place 1, 3, 7, and 12 weeks after surgery. However, this schedule was altered soon after study enrollment began. If the participant was scheduled for a surgeon’s office or physical therapy visit at the 3, 7, and 12-week time points, the surveys were administered in person. If participants’ follow-up visits were not scheduled for these time periods, then the SF-36 and KOS-ADLS surveys were mailed (with self-addressed stamped envelopes) in anticipation of these surveys being returned at the 3, 7, and 12-week points after surgery. These mailings were preceded and followed up with a reminder phone call. If the participant was not seen by the study coordinator at a scheduled office visit, then the participant was called, and a mailing to the participant’s home was arranged.
Data were first explored to determine demographic and day-of-surgery equivalence among treatment groups of patients included in the present analysis (Table 1). Demographic variables (gender, age, race/ethnicity, Body Mass Index (BMI), smoking history, surgeon, length of surgery and use of thigh tourniquet intraoperatively) were analyzed for differences between nerve block treatment groups using one-way ANOVA (for continuous variables) or the chi-square test (for categorical variables).
Both the SF-36 and the KOS-ADLS scores were administered on several occasions after surgery. Due to the variability in response times in returning the surveys to the research team, most participants had unequally spaced survey outcomes. Since the time intervals among surveys were not equal for most patients, we present the data in 2 ways. First, we present the descriptive data acquired according to the responses completed at 3, 7, and 12 weeks for all patients who returned completed surveys at any or all of these time intervals. For each time point with this method, the range of responses was tabulated from the specific week-point (i.e., 3, 7, and 12 weeks) plus and minus 5 days from each week-point. These descriptive data are presented adjacent (Figures 1–2) to the population norm for adults of age 25–34 years with no major medical (or surgical) problems.4,5 The age group of 25 to 34 years represents the 95% confidence interval of the mean age of participants in our study (Table 1). Second, we applied Generalized Estimating Equation (GEE) models to maximize the data analysis potential from all returned surveys (independent of survey return date relative to the day of surgery) and to evaluate the association among the 3 nerve block treatment groups and the SF-36 and KOS-ADLS outcomes. In other words, GEE modeling allowed us to accurately analyze all patient-reported outcome data (not just the data returned at the 3-, 7-, and 12-week time-points). We first used a univariate GEE model (with the only factor considered being the nerve block treatment group), followed by a multiple regression GEE model. In the multiple regression GEE models, we controlled for all other gathered covariate data, such as ASA/PS, race, gender, duration of surgery, smoking status, BMI, baseline SF-36 and KOS-ADLS scores. Again, the principal factor analyzed in the multiple regression GEE model was the nerve block treatment group.
It should be noted that the KOS-ADLS (but not the SF-36) was also administered to patients 1 week after surgery at the follow-up visit with the surgeon. These 1-week KOS-ADLS data are included in the multiple regression GEE equations for the KOS-ADLS.
P < 0.05 was considered statistically significant. All analyses were conducted using the Statistical Package for the Social Sciences (SPSS for Windows, version 14.0, Chicago, IL) and SAS (version 9.1.3, SAS Institute Inc., Cary, NC). In this current study of 7-day to 12-week patient-reported outcomes, our inclusion rate was 80%, which matched the 80% power/sample size threshold required for interpretation.1
Recruitment began in July 2001, and study follow-ups were completed by January 2005. Two-hundred-seventy patients consented to participate in the study. Thirty-five of the 270 recruited patients were excluded before the end of the day of surgery, with the reasons for exclusions having been detailed previously.1 Of the remaining 235 patients, 210 had complete baseline SF-36 data, and 217 had complete KOS-ADLS baseline data (i.e., no missing data entries). There were no demographic differences among treatment groups (Table 1) with respect to gender, age, race/ethnicity, BMI, ASA/PS, smoking history, use of thigh tourniquet or length of surgery. These 210 and 217 patients, respectively, had also at least 1 completed postoperative survey, yielding a total of 523 SF-36 surveys analyzed, and 745 KOS-ADLS surveys analyzed. The response rates (i.e., participants returning at least 1 completed survey during the postoperative course) on the SF-36 and KOS-ADLS did not significantly differ between nerve block treatment groups; the overall response rate was 99%.
Baseline data (n=217) for KOS-ADLS survey data were comprised of a mean total score of 69.96 (standard deviation = 19.94). There were no differences in baseline KOS-ADLS scores among nerve block treatment groups (data not shown).
The GEE results (Table 2) show that in both univariate and multiple regression models, nerve block treatment group was not significantly associated with SF-36 or KOS-ADLS outcomes (all with P ≥ 0.05). The multiple regression GEE model (Table 3) that considered covariates only (i.e., not evaluating the nerve block treatment group as a factor, but instead as a covariate) showed that both the PCS of the SF-36 (P < 0.0001) and the KOS-ADLS total score (P < 0.0001) increased over time after surgery. Baseline SF-36 PCS scores were a positive predictor of postoperative PCS scores, which was also true for the KOS-ADLS scores. Increasing patient age was a predictor of lower postoperative PCS scores, but age did not influence KOS-ADLS scores.
When the described spinal anesthetic (with multimodal analgesia) was used for patients undergoing ACLR, and when factoring the described perineural catheter treatment group, there were no statistically significant differences in general health status and knee function outcomes from 7 days to 12 weeks after surgery. More specifically, there were no differences in SF-36 or KOS-ADLS scores based on the nerve block treatment group considered in isolation (univariate GEE analysis) or when controlling for demographic covariates (multivariable GEE analysis). The same conclusion cannot be declared for ACLR patients undergoing any other anesthetic plan (e.g., general anesthesia [GA] with high-dose opioid analgesia, versus spinal anesthesia, multimodal analgesia, with or without a perineural femoral catheter). To reach this conclusion addressing GA-based anesthetic care plans, additional research would be needed.
It is clinically accepted that ACLR is not as invasive as total knee replacement. Therefore, it is logical that patients undergoing less invasive knee surgery may derive less long-term benefit after a continuous femoral perineural catheter than when compared with a more invasive surgery (such as total knee replacement with GA2,3 and the same perineural catheter). As described above, the hypothesis for ACLR patients needs to be re-tested in GA contexts that do not include spinal anesthesia and/or multimodal analgesia before drawing such conclusions.
Unlike inpatients undergoing total knee replacement 2,3 (and physical therapist-reported patient outcome in the hospital), patient-reported outcomes measures are a more appropriate methodology for outpatients. However, intermediate-term outcomes measures had, to that point, been unexplored in our subspecialty. To our knowledge, this is the first study that used validated general health and extremity function outcomes that measure the effects of nerve blocks (including continuous nerve block infusions with local anesthetics) on patient outcomes in the 7-day to 12-week period after surgery, in the setting of an anesthesia care protocol consisting of spinal anesthesia and multimodal analgesia.
The SF-36 has been used several times in previous publications to assess anesthesia-specific outcomes. Carli et al. showed that adding a thoracic epidural to a standardized anesthetic–analgesic–rehabilitation care plan for colonic surgery showed improved SF-36 outcomes at 3 and 6 weeks after surgery (when compared with patients only treated with IV patient-controlled opioids).11 In another study, Shyong et al. showed that the use of intraoperative and/or perioperative beta blockers led to improved SF-36 PCS scores (primarily through reduced bodily pain) 72 hours postoperatively, when compared with patients not receiving beta blockers (n=59, all patients 65 years or older).12
The 2 outcome studies in the previous paragraph (showing clinical differences in SF-36 outcomes based on perioperative anesthetic intervention) are offset by 2 reports of no difference in SF-36 outcomes. First, Wurm et al. showed no differences in SF-36 outcomes at 1 and 7 days postoperatively, but this study did not describe its basis of sample size determination (interscalene block performed preoperatively versus postoperatively, n=102).13 In a more recent study, no SF-36 differences were seen at 12 weeks postoperatively when comparing spinal anesthesia and GA for vaginal hysterectomy.14
A potential limitation of this study is that neither the SF-36 nor KOS-ADLS specifically addresses symptomatic outcomes associated with regional anesthesia procedures, such as peripheral nerve blocks (paresthesiae, or other symptoms associated with sensorimotor nerve damage). However, we recorded patients’ postoperative symptoms during this time on other case report forms, and all such symptoms (which were rare and generally not present at baseline) resolved by the 12-week period (usually much sooner). Such symptoms are considered rare, and the course of such complications typically resolves within a 6-month period.15
Our data show no outcome differences (i.e., 2-tailed) in the 7 days to 12 weeks after ACLR; therefore, there is no evidence to conclude that a femoral nerve block should be withheld based on adverse general health or knee function outcomes in the 7 days to 12 weeks after surgery. This finding serves as an addendum to our earlier report 1 that the femoral perineural catheter yields (i) a clinically significant improvement in numeric rating scores for pain with movement and (ii) lower opioid consumption, in the first 2 days after surgery.
The question then becomes, if such a labor-intensive intervention 16 yields a significant 2-day benefit and no long-term detriment (but no apparent long term benefit either), then is the intervention worthwhile? When considering pharmacoeconomic research addressing regional anesthesia and analgesia over the past decade, our patient care outcomes achieved in this protocol appear to warrant the labor-intensive intervention, when considering the context of the ambulatory surgery care process, described further below.
Patients ascribe value to the avoidance of nausea, vomiting and pain.17 The Joint Commission for the Accreditation of Healthcare Organizations mandates the evaluation of pain as the fifth vital sign 18 and implies that the avoidance of moderate-to-severe pain is an important patient care objective. The National Institutes of Health defines moderate to severe pain as “5 or higher” on a zero-to-10 scale.19 Since 75% of our study’s patients’ treatment infusion group (i.e., not the placebo infusion groups) accomplished this pain management-related public health objective 1, then the intervention appears to be worthwhile on that parameter. We have reported a low incidence of nausea and vomiting both on the day of surgery and through the first 4 days after surgery 20 in the context of regional anesthesia and multimodal analgesia-antiemesis, accomplishing described patient-oriented objectives both in-hospital and at home. After the day of surgery, the antiemetic benefits were not traceable to the perineural infusion, per se, but instead to the establishment and maintenance of analgesic and antiemetic “momentum,” in that the previous days’ symptoms were predictive of the next day’s symptoms when considered as covariates with the nerve block catheter treatment group. Therefore, the femoral perineural catheter may not necessarily provide additional value beyond spinal anesthesia and multimodal analgesia-antiemesis for the context of nausea and/or vomiting in hospital or after discharge.
From a hospital efficiency standpoint, hospital cost savings have been reported 21 when patients bypass the phase-1 postanesthesia care unit immediately after surgery 22, which is more reliably achieved after regional anesthesia with a sustained regional analgesic technique in place. 21,23–27 Therefore, our study’s patients with a perineural femoral catheter may not necessarily have additional value when compared with a simpler single-injection femoral nerve block. Standardization of patient throughput criteria has been recommended to capitalize on the hospital-based benefits of regional anesthesia 16,21,22,25–27, as well as to enhance the comparability of outcomes studies.16,28 Finally, the costs of avoided unplanned hospital admission 21 and pain-related hospital readmission 29 have been elucidated. One levobupivacine infusion patient in our study returned to the emergency department for extrapyramidal symptoms 30 and pain control after a hamstring autograft (the sciatic nerve was not blocked preoperatively per protocol); there were no other unplanned hospital admissions, readmissions or emergency room visits for pain control throughout our study. Therefore, the inconvenience of perineural catheter placement for the practitioner appears to be offset by the myriad health care system benefits, as well as patient benefits in the first few days after surgery, with no apparent patient-reported long-term detriment based on validated outcome surveys. We routinely administer sciatic single-injection nerve blocks for our ACLR patients undergoing hamstring autografts 31, but did not for this study.
One interesting finding in the multiple regression GEE model (Table 2) was that the LbLi patient group showed a trend (P=0.05) toward better scores on the MCS (i.e., mental component summary) than did the placebo SbSi group. Although this trend may reflect an artifact of multiple analyses, the finding may prove to be of value for future hypothesis generation, for example, when comparing more disparate anesthesia-analgesia care plans. Another interesting finding noted in Table 3, from the standpoint of epidemiology and public health, was that smoking status was associated with lower patient-reported SF-36 MCS scores during the 7-day to 12-week period after the described surgery and anesthetic.
One limitation of this study is that postoperative rehabilitation regimens, although standardized at the level of surgeons’ printed instructions to their surgical patients, could be monitored neither for individual therapist compliance to the written protocols, nor for patient compliance to the written protocols and/or the physical therapy objectives given by individual patients’ specific therapists. We do not foresee this methodologic barrier ever being overcome in study designs such as this with large patient samples. In fact, our experience early in the study was that requiring patients to undergo a centralized physical therapy program on site at the surgeons’ outpatient clinic and physical therapy center (as a condition for study inclusion) led to no patient recruitment for the first several months of study enrollment.
In conclusion, this triple-blinded, prospective, randomized, clinical trial addressed patient-reported general health status and knee function outcomes. For the period of 7 days to 12 weeks after ACLR surgery, these outcome data did not show any differences based on nerve block treatment group superimposed upon the described anesthetic-analgesic-antiemetic care plan. It is reassuring that, given the previously described 1 short-term benefits of continuous femoral perineural analgesia, this treatment had no patient-reported detriments during the described time frame, based on the parameters measured by 2 validated patient outcome surveys (SF-36, and KOS-ADLS).
Financial support (for Dr. B. Williams): National Institutes of Health/National Institute of Arthritis, Musculoskeletal, and Skin Diseases K23 AR47631, Bethesda, Maryland, United States; and International Anesthesia Research Society Clinical Scholar Research Award (2001), Cleveland, Ohio, United States. Additional support from a Seed Grant courtesy of the University of Pittsburgh Department of Anesthesiology, Pittsburgh, Pennsylvania, United States.
The lead author would like to acknowledge the teamwork provided by enrolling anesthesiologist Raymond B. Schwartz, MD (Assistant Professor, retired) from the University of Pittsburgh Department of Anesthesiology, Pittsburgh, Pennsylvania, United States. The authors also thank the surgeons from the University of Pittsburgh Department of Orthopaedic Surgery (Pittsburgh, Pennsylvania, United States), Center for Sports Medicine, who allowed us to enroll their patients: Freddie H. Fu, MD (Professor); Christopher D. Harner, MD (Professor); Robin V. West, MD (Assistant Professor), Patrick J. McMahon, MD (Assistant Professor); and Craig H. Bennett, MD (Assistant Professor). The authors also acknowledge previous research coordinators for this study based at the University of Pittsburgh: Chiara M. Figallo, MLIS (Department of Anesthesiology); and Kimberly A. Francis, MS, MPA (Department of Orthopaedic Surgery); and offer special thanks to the former Director of Orthopaedic Clinical Research (University of Pittsburgh), Molly T. Vogt, Ph.D., Dr.P.H.
Nerve stimulation needles (Prolong PL-50) were provided by Spinal Specialties, inc., San Antonio, Texas, United States; Life-TechR, inc., Stafford, Texas, United States; and I-Flow Corporation, Lake Forest, California, United States. Elastomeric nerve block infusion devices were provided by McKinley Medical, Wheat Ridge, Colorado, United States. Patient samples of rofecoxib were provided by Merck & Co., Inc., Whitehouse Station, New Jersey, United States.
Reprints will not be available from the author.