Article Text

Download PDFPDF

Peripheral nerve blocks are not associated with increased risk of perioperative peripheral nerve injury in a Veterans Affairs inpatient surgical population
  1. Meghana Yajnik1,
  2. Alex Kou1,2,
  3. Seshadri C Mudumbai1,2,
  4. Tessa L Walters1,2,
  5. Steven K Howard1,2,
  6. T Edward Kim1,2 and
  7. Edward R Mariano1,2
  1. 1 Department of Anesthesiology, Perioperative and Pain Medicine, Stanford University School of Medicine, Stanford, California, USA
  2. 2 Anesthesiology and Perioperative Care Service, Veterans Affairs Palo Alto Health Care System, Palo Alto, California, USA
  1. Correspondence to Edward R Mariano, Department of Anesthesiology and Perioperative Care Service, VA Palo Alto Health Care System, Palo Alto, CA 94304, USA; emariano{at}stanford.edu

Abstract

Background and objectives Perioperative peripheral nerve injury (PNI) is a known complication in patients undergoing surgery with or without regional anesthesia. The incidence of new PNI in a Veterans Affairs (VA) inpatient surgical population has not been previously described; therefore, the incidence, risk factors, and clinical course of new PNI in this cohort are unknown. We hypothesized that peripheral nerve blocks do not increase PNI incidence.

Methods We conducted a 5-year review of a Perioperative Surgical Home database including all consecutive surgical inpatients. The primary outcome was new PNI between groups that did or did not have peripheral nerve blockade. Potential confounders were first examined individually using logistic regression, and then included simultaneously together within a mixed-effects logistic regression model. Electronic records of patients with new PNI were reviewed for up to a year postoperatively.

Results The incidence of new PNI was 1.2% (114/9558 cases); 30 of 3380 patients with nerve block experienced new PNI (0.9%) compared with 84 of 6178 non-block patients (1.4%; p=0.053). General anesthesia alone, younger age, and American Society of Anesthesiologists physical status <3 were associated with higher incidence of new PNI. Patients who received transversus abdominis plane blocks had increased odds for PNI (OR, 3.20, 95% CI 1.34 to 7.63), but these cases correlated with minimally invasive general and urologic surgery. One hundred PNI cases had 1-year follow-up: 82% resolved by 3 months and only one patient did not recover in a year.

Conclusions The incidence of new perioperative PNI for VA surgical inpatients is 1.2% and the use of peripheral nerve blocks is not an independent risk factor.

  • regional anesthesia
  • peripheral nerve block
  • complications
  • nerve injury
  • patient safety

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Introduction

Perioperative peripheral nerve injury (PNI) is an uncommon but potentially serious complication after surgery, and reports continue to suggest regional anesthesia as an etiology.1 The reported incidence of perioperative PNI varies by surgery type2–4 and practice setting,3–5 but has not yet been described in a Veterans Affairs (VA) inpatient surgical population. Similarly, the risk factors for perioperative PNI in this population are unknown. VA patients tend to be older with greater chronic medical diseases and lower household incomes compared with the general population.6 7 Veteran status is an independent risk factor for homelessness,8 and approximately one in four veterans have at least one mental illness.9 The presence of a serious mental illness (eg, schizophrenia, bipolar disorder), substance use disorder, or depression in veterans increases the risk for hospitalization or death.9 In addition, veteran patients who have surgery at teaching VA hospitals have higher 30-day postoperative morbidity rates when compared with non-teaching VA hospitals.10 Therefore, previously described findings of perioperative PNI in non-VA medical centers may not be generalizable to the veteran population especially at university-affiliated VA hospitals.

We aimed to characterize the incidence, risk factors, and clinical course of new perioperative PNI in a cohort of surgical inpatients at a tertiary care university-affiliated VA hospital and hypothesized that the presence of a peripheral nerve block (NB) does not increase the incidence of PNI.

Methods

Model of perioperative care

A Perioperative Surgical Home (PSH) coordinated model of surgical care has been in place at a tertiary care university-affiliated VA hospital since 2010 with a custom electronic database since 2012, which has been previously described.11 12 This PSH database is designed for ongoing quality improvement and does not communicate with the electronic medical record. Within the PSH model at our institution, there is a dedicated Regional Anesthesiology and Acute Pain Medicine (RAAPM) team staffed by an anesthesiologist per day11 12 that performs NB procedures and directly manages postoperative pain for orthopedic surgery and neurosurgical spine surgery patients.11 Multimodal analgesic clinical pathways are managed by the RAAPM team, which include the routine inclusion of regional analgesic techniques for patients undergoing extremity, thoracic, breast, and minimally invasive abdominal and pelvic surgery. A different anesthesiologist is responsible for the patient’s intraoperative care. One PSH anesthesiologist is assigned out of the operating room per day, apart from the RAAPM attending, to provide physician oversight of the preanesthetic evaluation clinic, specialty consultation for preoperative optimization of medically complex patients, and routine postoperative follow-up of every patient who has received anesthesia the previous day using a standard checklist.11 The PSH anesthesiologist continues to follow inpatients with active issues beyond postoperative day (POD) 1; these issues may include, but are not limited to, continued mechanical ventilation or initiation of non-invasive positive airway pressure, hemodynamic instability, persistent anemia, new arrhythmia, and new PNI. Within the PSH the anesthesiologist’s daily workflow is entry of data for POD 1 follow-up patients, POD 30 outcomes data from chart review of patients who underwent surgery and anesthesia 30 days prior, and patient and procedural data for the next day’s scheduled cases into the PSH database.

Data elements

We conducted a 5-year review of the PSH database to evaluate all consecutive surgical patients admitted to the hospital postoperatively for incidence of new PNI. The assessment of new PNI is part of the POD 1 follow-up evaluation by the PSH anesthesiologist, and this is explicitly differentiated from worsening of pre-existing neuropathy. We use a standard script when investigating perioperative PNI, which starts with a question ‘Since surgery, have you noticed any areas on your body of new numbness, tingling, or weakness?’ If the response is affirmative, the site is examined to identify a potential nerve distribution, and this information is included in the PSH database. All patients with peripheral NBs, single-injection and continuous, are routinely seen POD 1 by the PSH attending and also followed daily by the RAAPM team until block resolution. Other routine POD 1 data fields include level of pain control, postoperative nausea and vomiting, sore throat, satisfaction with perioperative care, communication with primary team or other specialists, transfer to higher level of care, and less common postoperative complications (eg, corneal abrasion, postdural puncture headache). Baseline characteristics collected for each patient include age, American Society of Anesthesiologists (ASA) physical status, type of surgery (categorized into surgical subspecialty, although a detailed description of the actual surgical procedure is available for each case), anesthesiologist and surgeon assigned to the case, anesthetic technique, and regional anesthesia details (eg, site, catheter vs single-injection) when applicable.

Outcomes

The primary outcome was incidence of new PNI between groups that did or did not have NB. Secondary outcomes included rates of PNI within groups based on patient baseline characteristics, anesthetic technique, type of NB, and surgery type. Individual electronic charts of patients with new PNI were manually reviewed to characterize the clinical course for up to a year following surgery.

Statistical analysis

Experimental analyses were conducted using R V.3.3.2 statistical programming language (R Foundation for Statistical Computing, Vienna, Austria).13 Potential confounding factors were first examined individually against the incidence of neuropathy, included as independent variables in a logistic regression model, and finally as part of a mixed-effects logistic regression model.14 Model specification and fit were guided both by interpretability and statistical evaluation (ie, Akaike’s Information Criterion). Initially, independent variables including patient age, ASA status, emergency, and anesthetic technique were forced into the model with identity of primary surgeon entered as random effect. Given univariate analyses showing a significant association between administration of transversus abdominis plane (TAP) block and our primary outcome, the final model included TAP block as a binary variable. Model discrimination was evaluated by calculating the area under the receiver operating characteristic curve (AUC). All statistical tests were two-sided, with a p value <0.05 considered statistically significant.

Results

All 9558 cases in the database from November 1, 2012 to November 1, 2017 were included in the analyses. Over one-third of these cases received NB (n=3380). Overall, there were 114 cases of new perioperative PNI across eight general surgical categories (figure 1), for an incidence of 1.2%.

Figure 1

Distribution of new perioperative peripheral nerve injury cases within the inpatient surgical population (%) by surgical category (n=114 total).

Incidence of new PNI with and without NB

For the primary outcome, 30 of the 3380 patients who received NB experienced new PNI (0.9%) compared with 84 of the 6178 patients who did not receive NB (1.4%; p=0.053).

Incidence of new PNI based on patient and anesthetic factors

The incidence of new PNI for patients who received general anesthesia (GA) was 102 of 7398 (1.4%) vs 12 of 2160 (0.6%) for patients who did not receive GA (p=0.003). The age (mean (SD)) of patients with new PNI was 62 (11) years vs 65 (11) years for patients who did not develop PNI (p=0.001). Patients classified as ASA 3 had a lower incidence of PNI (70/6266) compared with ASA 1 or 2 patients (21/906; p=0.011). Emergent case status and non-emergent case status were not different in terms of new PNI incidence (6/822 vs 108/8736, respectively; p=0.267). The ORs of peripheral nerve injury occurrence for anesthetic-related factors based on the final mixed-effects logistic regression model are shown in table 1. The AUC for the final model was 0.76, indicating a good to strong model.

Table 1

ORs of peripheral nerve injury occurrence for anesthetic factors based on the final model

Incidence of new PNI based on the type of block

Although the overall incidence of new PNI was not affected by the presence of any NB (primary outcome), 13 of 2242 patients with a perineural catheter in the NB group experienced new PNI (0.6%) compared with 17 of 1084 single-injection NB patients (1.6%; p=0.017). Of these 30 cases, TAP single-injection or catheter accounted for 13 (43.3%; figure 2); 11 patients (84.6%) complained of new sensory and/or motor changes in their hands, and 10 of these 11 patients (90.9%) underwent minimally invasive general or urologic surgery. Only 8 of the 30 (26.7%) new PNI cases in patients who received NB were considered potentially related to the block. In the mixed-effects logistic regression model, patients who received TAP blocks had three times increased odds of developing PNI (OR 3.20, 95% CI 1.34 to 7.63), and these cases correlated with general surgery and urology procedures in combination with GA.

Figure 2

Distribution of new perioperative peripheral nerve injury cases within the subset of surgical inpatients who received nerve block (n=30 total) by nerve block type among patients who received a nerve block. TAP, transversus abdominis plane.

Clinical course of patients with PNI

Further characteristics of those with new perioperative PNI can be found in table 2. Three of 114 PNI cases were lost to follow-up. One hundred patients (88%) with PNI had completed 1 year of follow-up at the time of the study: 82% resolved by 3 months, 84% by 6 months, 86% by 9 months, and 87% by 12 months. Eleven patients experienced partial recovery and one patient did not recover in a year. Of the patients with new nerve injury, 87 (76%) had sensory neuropathy, while 24 (21%) had mixed sensorimotor neuropathy. The majority of patients with new neuropathy did not receive specialty care through a consult (85%). When a consult was placed, the most common specialty consulted was neurology (77%), followed by physical medicine and rehabilitation (15%), and other surgical specialties (7%). Although most patients did not receive any new prescribed therapy for their nerve injury, the most common prescribed treatment was physical therapy (9.6%). Only 3% of patients were prescribed new oral medication in response to PNI.

Table 2

Clinical course for patients with new postoperative peripheral nerve injury

Discussion

This is the first study to quantify the incidence of perioperative nerve injury and describe the clinical characteristics of PNI within a VA inpatient surgical population. Our data suggest that new perioperative PNI is not a rare finding among our VA patients and tends to occur in younger and healthier patients. The incidence of new perioperative PNI within our cohort may not differ significantly from similar single-site studies involving other inpatient surgical populations.2–5 While many studies have reported the risk of PNI within cohorts of patients who all receive a regional anesthetic,15–17 the perspective provided by including surgical patients in the denominator who do not receive regional anesthesia is typically lacking and should be included in the discussion of perioperative PNI.18 Having the PSH in place, with prospective data collection on perioperative outcomes not routinely included in electronic health records, has led to much more attentive care and has fostered a culture of continuous improvement in our practice. Based on the present study involving a large cohort of consecutive surgical inpatients, the risk of new perioperative PNI is not increased by peripheral nerve blockade even in a medically complex VA patient population that may be susceptible to complications.

The new perioperative PNI incidence of 1.2% that we report is much higher than the 0.03% previously reported from a large retrospective cohort study of over 380 000 anesthetized patients by Welch and colleagues.19 This study included inpatient and ambulatory surgical cases and used three sources of data: a voluntary reporting quality assurance database, closed claims, and billing data.19 In contrast, our study relies on a single source of prospectively collected data for surgical inpatients only, and thus complications are less likely to be under-reported. The PSH model of care employed at our institution11 differs from other anesthesiology practices and even other PSH models,20 and our ability to detect new perioperative PNI immediately after surgery results from standard work built into this model. All patients who receive anesthesia and are admitted to the hospital after surgery are routinely screened for neuropathy on their first POD as well as other perioperative outcomes of interest (eg, corneal abrasion, sore throat from tracheal intubation, and postoperative nausea and vomiting) that are not typically collected by nurses or other healthcare professionals, and data are prospectively entered by the PSH attending physician, who ensures data validity. Having a PSH is a culture change. By integrating active surveillance into clinical practice through the routine collection of predefined perioperative outcome measures, the PSH regularly inspires continuous improvement activities across multiple disciplines involved in the care of the surgical patient.

The present study adds further evidence to support previous authors’ conclusions that the use of peripheral nerve blockade is not an independent risk factor for developing new perioperative PNI.2–4 In fact, patients who avoided GA had a lower incidence of PNI compared with patients who received GA, and supplementing GA with any regional anesthesia technique decreased the odds of PNI by 61%. However, this finding does not exonerate the regional anesthesiologist of responsibility, and there may be block-related factors that do deserve consideration, such as mechanical trauma, local anesthetic neurotoxicity, and nerve ischemia from vasoconstriction.21 In most cases of perioperative PNI, the etiology will be multifactorial, with patient comorbidities and surgical factors as important contributors.18 21 It is important to remember that association and causation are not the same, and the association between TAP block and new PNI in our study is a good example. Taken at face value, TAP blocks appear to increase the risk of PNI; however, none of the cases were reasonably block-related, and the combination of hand neuropathy and minimally invasive surgery suggests patient positioning as a probable etiology. This finding would not have been possible to quantify without the PSH model of care and PSH database employed at our institution. Issues related to perioperative PNI and regional anesthesia continue to make up a large proportion of anesthesia malpractice claims in non-spine orthopedic surgery,1 and expecting patients to call if something bad happens does not represent adequate follow-up. Perhaps a suggested best practice is consistent proactive follow-up to monitor postoperative outcomes for all surgical patients regardless of the type of anesthesia administered.

When patients do experience new perioperative PNI, it is important to provide appropriate, evidence-based, and individualized treatment.22 Routine electrophysiologic testing immediately after injury is controversial and likely will not change treatment.22 In our cohort, more than 80% resolved completely by 3 months, and a minority received subspecialty consultation or a newly prescribed treatment. Unfortunately, when long-lasting PNI does occur, it may have serious consequences. One case of nerve injury has been reported by a VA hospital prior to our study, and it involved a patient who received an interscalene brachial plexus block for a rotator cuff repair, then suffered a prolonged hemidiaphragmatic paralysis in association with three cranial nerve palsies.11 The only PNI case in our study that did not have any degree of recovery was a lateral femoral cutaneous nerve injury in a patient who underwent anterior cervical discectomy and fusion.

There were several important limitations to our study. First, the PSH database only includes data from surgical inpatients. Therefore, we cannot comment on the risk of new perioperative PNI in ambulatory surgical patients. Second, there is an imbalance in the size of each cohort as can be expected in a retrospective cohort study. We included all consecutive surgical inpatients in order to minimize selection bias, and there were no changes in clinical pathways with respect to peripheral nerve blockade during the study time period. Third, the PSH database does not communicate with the VA electronic medical record and lacks detailed patient-level comorbidity information. Previous studies have shown that patient factors such as pre-existing neuropathy and diabetes may play an important role in the etiology of PNI.19 21 22 The standard work for the PSH anesthesiologist requires differentiation of new PNI from exacerbation of pre-existing neuropathy during the bedside evaluation through patient history and symptom report, and new PNI is separated from pre-existing neuropathy based on these factors in the database. Patients at our institution do not receive routine preoperative electrophysiologic testing before elective surgery for comparison with possible postsurgical changes, and our perioperative PNI protocol does not mandate early electrophysiologic testing unless indicated by severity of presentation and/or inability to participate in physical therapy. Therefore, our rate of new PNI may be an overestimate, although it falls within the ranges reported by other studies previously. Although this potential misclassification of chronic neuropathy exacerbation as new PNI is a limitation, we believe that erring on the side of overdiagnosis (ie, ‘assuming the worst’) is not necessarily a bad thing when it comes to managing patients with perioperative PNI. Fourth, although all patients were examined on POD 1 and followed until resolution of active issues, patients with delayed presentation of nerve injury may have been missed. Finally, the results of this single-center study may only apply to patient cohorts with similar characteristics and surgical/anesthetic exposures and should not be generalized to all surgical patients.

In conclusion, the incidence of new perioperative PNI for surgical inpatients at a tertiary care VA hospital is 1.2%, and peripheral nerve blockade is not an independent risk factor. Younger and healthier surgical patients may be at higher risk, as well as those who undergo minimally invasive abdominal and pelvic surgery. While most cases of new PNI resolve within 1 year, anesthesiologists must remain vigilant and develop systems such as PSH to routinely monitor for new perioperative PNI in all surgical patients.

References

Footnotes

  • Presented at This work was presented in part at the 2018 World Congress on Regional Anesthesia and Acute Pain Medicine Meeting, April 19–21, 2018, New York, New York.

  • Twitter New study by @EMARIANOMD: "Peripheral Nerve Blocks Are Not Associated with Increased Risk of Perioperative Peripheral Nerve Injury in a Veterans Affairs Inpatient Surgical Population"

  • Competing interests None of the authors has any conflicts of interest to declare.

  • Ethics approval This retrospective cohort study was reviewed and approved with waiver of informed consent by the local institutional review board (Stanford, California) and Veterans Affairs (VA) Research Committee (Palo Alto, California).

  • Provenance and peer review Not Commissioned; Externally peer reviewed.