Can bedside patient-reported numbness predict postoperative ambulation ability for total knee arthroplasty patients with nerve block catheters?

Abstract

Background

Adductor canal catheters offer advantages over femoral nerve catheters for knee replacement patients because they produce less quadriceps muscle weakness; however, applying adductor canal catheters in bedside clinical practice remains challenging. There is currently no patient-reported outcome that accurately predicts patients' physical function after knee replacement. The present study evaluates the validity of a relatively new patient-reported outcome, i.e., a numbness score obtained using a numeric rating scale, and assesses its predictive value on postoperative ambulation.

Methods

We conducted a retrospective cohort study pooling data from two previously-published clinical trials using identical research methodologies. Both studies recruited patients undergoing knee replacement; one studied adductor canal catheters while the other studied femoral nerve catheters. Our primary outcome was patient-reported numbness scores on postoperative day 1. We also examined postoperative day 1 ambulation distance and its association with postoperative numbness using linear regression, adjusting for age, body mass index, and physical status.

Results

Data from 94 subjects were included (femoral subjects, n = 46; adductor canal subjects, n = 48). Adductor canal patients reported decreased numbness (median [10^th–90^th percentiles]) compared to femoral patients (0 [0–5] vs. 4 [0–10], P = 0.001). Adductor canal patients also ambulated seven times further on postoperative day 1 relative to femoral patients. There was a significant association between postoperative day 1 total ambulation distance and numbness (Beta = –2.6; 95% CI: –4.5, –0.8, P = 0.01) with R² = 0.1.

Conclusions

Adductor canal catheters facilitate improved early ambulation and produce less patient-reported numbness after knee replacement, but the correlation between these two variables is weak.

Introduction

For patients undergoing total knee arthroplasty (TKA), evidence supports the use of continuous peripheral nerve block (CPNB) techniques with femoral or saphenous perineural local anesthetic infusions as part of a perioperative multimodal analgesic regimen [1,2,3]. Although CPNB offers clear analgesic advantages with selectivity for the operative extremity by targeting specific nerve distributions, there is rising concern over the quadriceps weakness produced by femoral nerve blocks and the potential risk of falls [4,5]. More recently, the adductor canal catheter (ACC) technique [6] has emerged as an alternative to the femoral nerve catheter (FNC) for TKA patients, because it better preserves quadriceps motor function while still providing effective pain relief [7,8], and may facilitate improved early ambulation after surgery [9].

While previous studies have assessed quadriceps strength in terms of maximum voluntary isometric contraction [10,11,12], this method requires the use of a dynamometer and is not routinely part of clinical bedside assessment. Clinical assessments commonly rely on physical examination and subjective patient-reported outcomes (e.g., numeric ratings of pain). Although the use of numeric rating scales (NRS) of patient-reported numbness has been reported previously [13,14], this measure has not yet been correlated with physical function and also remains to be evaluated as a predictor of postoperative ambulation. A patient who self-reports numbness of the quadriceps muscle on the operative extremity after TKA may be more reluctant to participate in physical therapy due to fear of falling. Given the expected increase in TKA prevalence over the next 10 to 15 years, due to an aging population and rising prevalence of osteoarthritis [15], there is a need to critically evaluate assessment techniques and incorporate patient-reported outcomes as appropriate.

We designed this study to test the hypothesis that TKA patients with an ACC will exhibit lower self-report numbness scores on an NRS compared to patients with an FNC. We also explored the potential correlation between patient-reported numbness and ambulation distance on postoperative day (POD) 1.

Materials and Methods

After receiving Institutional Review Board approval with waiver of informed consent (Stanford University School of Medicine, Stanford, CA, USA), we conducted this retrospective cohort study pooling data from two previously-published randomized clinical trials involving TKA patients [13,14]. These two sub-studies were part of a larger prospective project registered under NCT01459523 (http://clinicaltrials.gov); one compared proximal and distal ACC [14], and the other compared long-axis and short-axis ultrasound imaging for in-plane FNC insertion [13].

Original study criteria and outcomes

The inclusion and exclusion criteria were the same for both studies. Adults (≥ 18 years of age) scheduled for unilateral knee arthroplasty with a perineural catheter planned for postoperative analgesia were assessed for inclusion. We excluded patients unable to understand the study protocol or care for the infusion pump/catheter system, and those with any known contraindication to the study medications or regional anesthesia, insulin-dependent diabetes mellitus, neuropathy of any etiology in the affected extremity, hepatic or renal failure, any additional surgical site outside of the limb intended for catheter placement, chronic opioid use, active illicit substance abuse, pregnancy, or inability to communicate with the investigators and hospital staff.

Preoperatively, all patients underwent perineural catheter insertion (Arrow FlexTip Plus, Research Triangle Park, NC, USA) under ultrasound guidance at either the femoral nerve or adductor canal location. ACCs were placed adjacent to the saphenous nerve lateral to the superficial femoral artery in the subsartorial tunnel just distal to the apex of the femoral triangle (considered "proximal" [14]) or at the mid-thigh, halfway between the anterior superior iliac spine and the top of the patella as described by Lund et al. [6].

The intraoperative anesthesia technique has not been strictly standardized, but all subjects enrolled in the previous studies [13,14] received general anesthesia, combining inhalational and intravenous agents, without additional neuraxial block. All patients received the following standardized perioperative multimodal analgesic regimen designed for TKA: surgeon-administered periarticular injections of ropivacaine 0.2% (150 ml) and epinephrine (2.5 µg/ml), with ketorolac (30 mg) divided equally within the posterior capsule, retinacular layer, and subcutaneous tissue at the conclusion of surgery [16,17]; perineural infusion of ropivacaine 0.2% (basal rate of 6 ml/h; patient-controlled bolus of 5 ml; 30-minute lockout interval); and scheduled oral oxycodone and acetaminophen tablets (5 and 325 mg, respectively) four times daily, oral diclofenac (50 mg) twice daily, and additional oral oxycodone (5–10 mg) and intravenous (IV) nurseadministered opioid (morphine [1–2 mg] or hydromorphone [0.2–0.4 mg]) as needed every 2–3 hours for breakthrough postoperative pain inadequately treated with the perineural ropivacaine infusion/bolus and oral agents.

The only difference among groups was the location of the perineural catheter. The primary outcome of these original studies [13,14] was the onset time of anesthetic block in minutes. Other measurements collected on POD 1 included: average and worst post-surgical pain (NRS, 0–10), opioids consumed (mg oxycodone and/or IV morphine), number of awakenings due to pain, satisfaction with pain control (measured on a Likert scale; 0 = not at all satisfied, 10 = extremely satisfied), leakage of fluid from the catheter site, and subjective global rating of numbness (lack of sensation) in the operative extremity (0 = normal sensation; 10 = completely insensate).

The primary outcome of the present study was patient-reported numbness score on POD 1. Secondary outcomes included total daily ambulation distance (meters) for POD 1, in addition to other measures collected at the time of the original studies mentioned previously. All data from all subjects who completed the study were included in the analyses. Descriptive statistics were performed, and normality was assessed by the Kolmogorov-Smirnov test. For normally-distributed data, single comparisons between the ACC and FNC groups were performed using Student's t-test; for non-normal continuous data, the Mann-Whitney U test was used. The Z test or Fisher's exact test were used for comparisons of categorical data. As a secondary analysis, we examined the association between total daily ambulation distance and numbness for POD 1 using ordinary least-squares linear regression, adjusting for age, body mass index, and ASA status. NCSS-PASS (Kaysville, UT, USA) and STATA (ver. 12.1; STATA Corp., College Station, TX, USA) software packages were used for analyses. All P values were two-sided, and P < 0.05 was considered statistically significant.

Results

The data from all 94 previous clinical trial subjects were included in this study (46 FNC and 48 ACC). The FNC group was comparable to the adductor canal group in key baseline criteria (Table 1) although we did note a higher proportion of women in the femoral group relative to the ACC group (4/42 vs. 0/48, P = 0.04).

Table 1. Morphometric Measures and Administrative Data.

	Femoral (n = 46)	Adductor canal (n = 48)	P value
Age (yr)	65 (55–82)	66 (58–76)	0.34
Female/male (n)	4/42	0/48	0.04
ASA physical status	3 (2–3)	3 (2–3)	0.52
Height (cm)	178 (163–188)	178(168–185)	0.68
Weight (kg)	103 (77–130)	101 (78–135)	0.24
BMI (kg/m2)	32 (27–44)	32 (26–41)	0.90
Type of surgery (n)			0.99
Total knee arthroplasty	45	46
Unicompartmental knee arthroplasty	1	2

Values are reported as medians (10^th–90^th percentiles) or numbers of subjects (n), as indicated; ASA: American Society of Anesthesiologists, BMI: body mass index.

ACC patients reported lower numbness scores relative to the FNC group (median [10^th–90^th percentiles]): 0 (0–5) vs. 4 (0–10), P < 0.01. The ACC group also demonstrated an approximately seven-fold increase in ambulation on POD 1 relative to the FNC group (median [10^th–90^th percentiles]): 45.7 (6.2–91.4) vs 6.1 (0–42.7), P < 0.01. No differences were noted between the two groups on our other, secondary outcomes (Table 2). While linear models showed a statistically significant association between total daily ambulation distance and numbness on POD 1 (Beta = –2.6; 95% CI: –4.5, –0.8, P < 0.01), the R-squared was 0.1.

Table 2. Opioid Consumption and Patient-reported Measures on Postoperative Day 1.

	Femoral (n = 46)	Adductor canal (n = 48)	P value
Opioid consumption (mg)	1.3 [4.8]	1.5 [3.6]	0.89
Average pain (NRS 0–10)	3 (0–7)	4 (0–8)	0.34
Worst pain (NRS 0–10)	6 (2–10)	7 (0–10)	0.67
Satisfaction (Likert 0–10)	10 (8–10)	10 (8–10)	0.97
Awakenings due to pain (n)	3.0 [3.8]	1.9 [3.5]	0.17

Values are reported as medians (10^th–90^th percentiles) or means [standard deviation], as indicated, NRS: numeric rating scale (0–10; 0 = no pain, 10 = worst pain), n: number of events.

Discussion

For TKA patients receiving perineural local anesthetic infusions as part of a multimodal analgesic regimen, ACC results in less self-reported numbness and facilitates improved early ambulation when compared to FNC with no differences in pain control. Few other studies involving perineural catheters have focused on numbness as a procedure-specific outcome after surgery [18,19,20]. While self-reported measures such as the Western Ontario McMaster Osteoarthritis Index exist and are commonly used for long-term postoperative evaluation, these measures tend to be global, incorporating multiple dimensions such as overall well-being and independence [21,22] and have not been correlated with actual physical function. To better assess the role of regional analgesia, researchers have advocated for more clinically relevant patient-reported outcome measurement tools applicable to TKA patients [23].

Perineural catheter location is an important factor determining patient-reported numbness beyond type of local anesthetic, dosing strategy, and duration of local anesthetic infusion [24]. The 4-point median difference in numbness between the ACC and FNC groups is notable given that half of the ACC patients reported no numbness, and ACC has been previously shown to better preserve quadriceps muscle strength [11,12]. Patient-reported outcomes related to TKA have focused primarily on pain intensity [23], but the present study provides evidence that self-reported numbness using NRS may be easily incorporated into the bedside assessment to provide additional information. This is a clinically important outcome because the patients' subjective feeling of numbness may decrease their willingness to participate in physical therapy; physiologically, loss of proprioception and motor weakness contribute to gait and balance problems that can affect ambulation [25].

However, in our secondary analysis, although there was a significant association between total ambulation distance and numbness, 90% of the variance remained unexplained; therefore, self-reported numbness cannot be used to predict ambulation achievement. The remaining unexplained variance emphasizes the need for further research into the factors that potentially contribute to postoperative ambulation after TKA. These results are consistent with prior reports suggesting that patient-reported outcomes correlate poorly with performance-based outcomes [23,26]. Furthermore, we found no group differences in satisfaction, pain experience, or opioid consumption. Since patients' own perceptions regarding their functional ability and physical status may differ greatly from their actual performance outcomes when measured, our data reinforce the need for both performance-based and self-reported outcomes when assessing the progress of TKA patients' postoperative rehabilitation [27].

Our study had several limitations that should be considered when interpreting the results. First, we used a retrospective study design such that the reported associations may not be true causal relationships; our study also pooled data from two separate clinical trials that were not powered to test for the present study's primary outcome, even though both studies measured it. We believe that the fact that we combined data from two studies may explain the difference in number of female patients per group, because randomization was not applied across the entire sample. However, a potential advantage of our study is that, although it was retrospective, it used prospectively collected data during the immediate postoperative period when the effects of the ACC and FNC techniques on numbness would be most likely to be found using the same standardized study methodology and case report forms. Although selection bias remains possible, all patients were treated on the same clinical pathway for TKA, in which only the placement location of CPNB varied. Second, the generalizability of the study results is limited to practices employing similar techniques and equipment to those used in this study, and may not be generalizable to other institutions or non-veteran populations.

In summary, although perineural local anesthetic infusion in the adductor canal results in improved early ambulation distance and lower patient-reported numbness after TKA compared to femoral perineural infusion, the correlation between these two variables was weak. Future studies evaluating the comparative effectiveness of regional analgesic techniques in TKA patients should include both patient-reported and performance-based outcome measurements.