Chapter 74. Pharmacoeconomics of Regional Anesthesia: Implications in Ambulatory Orthopedic Surgery, Hospital Admission, & Early Rehabilitation

The analysis of the relationship of pharmaceutical and device costs to health care systems has been termed pharmacoeconomics, and four types of analytical techniques are commonly used for this purpose, namely cost-minimization, cost-benefit, cost-effectiveness, and cost-utility analyses.1

With diminishing payments from private and government-based health insurance programs worldwide, physicians and administrators are forced to focus attention toward cost-containment in order to maintain a profitable (or at least “break-even”) enterprise. Cost analysis is an emerging tool in health care economics, which can help physicians and administrators meet these new challenges.

Cost analysis examines health care expenditures, and the subtypes of cost analysis also examine factors that are inserted into a denominator of a cost equation. Such factors include monetary benefits (eg, cost-benefit analysis, or CBA), incremental changes of health-status variables (eg, cost-effectiveness analysis, or CEA), and patient-reported quality of life (eg, cost-utility analysis, or CUA).2 If outcomes are determined to be equivalent regardless of the treatment program implemented, then a basic cost-minimization analysis is all that is required, since the denominators are equal and the only relevant comparison is between the cost numerators of the compared programs.3

CEA is applicable when the effects of comparable health treatments or services share the same therapeutic goals, but have different degrees of effectiveness.4 With CEA, the analyst can compare alternative treatment strategies so that results can be expressed in identical effectiveness units. CEA accounts for the effect of a treatment plan on all clinical outcomes and its economic implications, rather than considering only the cost of devices, supplies, and pharmaceuticals.4 Effectiveness indicators, such as the number of adverse effects avoided or hospital stay reductions, are useful for comparing the different therapeutic alternatives considered. For this reason, CEA is one way of comparing treatment plans with the same desired effect but different outcome profiles, thus producing results expressed in terms of the number of adverse effects avoided. This approach implies weighting all adverse effects alike, or weighting the different adverse effects in the way deemed most suitable by the analyst.4

With respect to anesthesia selection, it is highly unlikely that comparing regional (RA) with general anesthesia/volatile agent (GAVA) techniques would show equal benefits or equal effectiveness (ie, life-years gained, days of disability avoided). RA significantly differs from GAVA, and the relevant side effect profiles and risks are quite different as well. In fact, in the past few years, it has become very clear in ambulatory procedures, for example, that the choices of the anesthetic and postoperative analgesic techniques have significant consequences on both the length of hospital stay and the frequency of unplanned hospitalization,5,6 and consequently, the overall cost of the surgery. As a result, comparisons between RA and GAVA would require a cost-benefit, cost-utility, or cost-effectiveness analysis.

As physicians caring for individual patients, it is important to note that basing a clinical practice strictly on cost analysis is not well advised, since individual patients have individual needs. However, in the setting of identifiable patient and hospital benefits when RA is used instead of GAVA, proper patient education regarding the benefits of RA is a most necessary step before any patient or health system benefits can be enjoyed.

It is also important to note that the literature is not replete with conclusive evidence addressing cost-effectiveness of RA versus GAVA for surgical procedures in which both options are viable alternatives for patient care. Limited studies available can be used to apply cost analysis estimations for other studies that did not specifically address costs, but such approaches must be interpreted with caution.

Selection of Techniques, Drugs, & Agents

Rationale for Avoiding General Anesthesia with Volatile Agents in Outpatients

When considering invasive outpatient orthopedic surgery, the routine use of GAVA without peripheral nerve blocks as the centerpiece of a multimodal analgesic plan is commonly associated with the following costly outcomes: (1) postanesthesia care unit (PACU) admission; (2) multiple nursing interventions for pain and postoperative nausea or vomiting (PONV); (3) PACU and same-day surgery discharge delays; and (4) unplanned hospital admission.7–11 However, most hospital pharmacy and therapeutics committees are primarily focused on the pharmacy budget. Six percent or less of all hospital costs related to surgical care are attributed to pharmacy drug costs,12 and examining drug costs in isolation without regard to patient outcomes is ill-advised. The “least expensive” outpatient GAVA technique, from the standpoint of drug acquisition for line items used for anesthesia induction and maintenance, would probably include thiopental, succinylcholine, opioids and volatile agents.

Postoperative Nausea & Vomiting & Comprehensive Multimodal Analgesia in Outpatient Regional Anesthesia Practice

A traditional GAVA plan (thiopental, succinylcholine, volatile agents, and opioids) is traditionally favored for these specific line-item budgets of the anesthesia department and the hospital pharmacy, but the technique is fraught with “downstream” expenses for the hospital and presents a common basis for patient dissatisfaction. It has been well documented that GA techniques that exclusively use propofol for induction and maintenance (or as an intravenous sedative technique combined with RA) has a significantly lower PONV rate than do GAVA techniques,10,13,14 and volatile agents and opioids are considered to be the leading causes of PONV.15 Exclusive use of propofol instead of volatile agents is considered to provide an important but incomplete method of prevention of PONV.15 Three antiemetics of differing mechanisms are required to prevent equivalent PONV outcomes in GAVA patients versus propofol total intravenous technique patients.16 RA techniques (without GA) are historically considered to be protective against PONV,17 but recent research has indicated that there can be significant differences in PONV incidence when differing RA techniques are compared for the same type of surgical procedure.18 In fact, patients undergoing single-injection peripheral nerve block (PNB) techniques that are later prescribed intravenous patient-controlled analgesia (IVPCA) do not appear to have significant benefit of PONV prevention versus similar patients undergoing GAVA with IVPCA,17 whereas patients receiving sustained analgesia primarily with continuous PNBs obtain additional PONV prevention benefits.17

For simple knee arthroscopy (mean procedure time <30 min), GA with desflurane was compared with ipsilateral hyperbaric bupivacaine spinal anesthesia (4 mg) to assess PONV and a myriad of other outcomes.19 Desflurane patients received PONV prophylaxis with dexamethasone and ondansetron if at least two risk factors were present (female, nonsmoker, PONV history, or motion sickness). Despite PONV prophylaxis in at-risk desflurane patients, all desflurane patients encountered significantly more PONV (6/32, 19%) than did selective spinal anesthesia patients (zero PONV, p = 0.024).19 In addition, desflurane patients encountered significantly more pain in both phases of postoperative recovery and more “extreme tiredness” in phase II recovery than did patients receiving selective spinal anesthesia.19

Generally speaking, for outpatient orthopedic anesthesia routinely incorporating RA techniques: (1) GAVA should be avoided; (2) opioids should be minimized; and (3) at least two if not three antiemetics of differing mechanisms of antiemetic prophylaxis should be used. My suggested criteria for multimodal antiemetics in outpatient orthopedics are: (1) if the surgery duration is greater than 30 min; (2) if it is likely that opioids will be included in the postoperative anesthesia plan; (3) if intraarticular analgesic adjuncts are used that may increase PONV risk (eg, opioids, neostigmine); or (4) if GAVA is to be used. In outpatient orthopedics, use of a volatile agent alone is a sufficient sole risk factor for routine multimodal antiemetic prophylaxis, regardless of other risk factors. I suggest (1) oral perphenazine18,20,21 (8 mg preoperatively, with consideration of a repeat 4-mg oral dose postoperatively should breakthrough PONV symptoms occur, but avoided in patients with Parkinson disease or history of adverse extrapyramidal reactions to phenothiazines); (2) dexamethasone18,22–26 (4 mg intravenously, perhaps avoided in diabetic patients); and (3) a 5-HT3 serotoninergic antagonist. The combination of dexamethasone and a 5-HT3 antagonist is more effective in preventing nausea and vomiting after discharge than is a 5-HT3 antagonist used alone.27

It should be understood that opioids are commonly going to be used for RA procedure premedication, intraoperative management of nerve block onset latency, and postoperative analgesic rescue of symptomatic pain not covered by the peripheral block's nerve distribution. In addition, multimodal analgesic techniques, such as low-dose intravenous ketamine,28–31 pre- and postoperative inhibitors of the cyclooxygenase type-2 enzyme,32–35 and intraarticular injections by the surgeon to cover myriad mechanisms of the acute pain inflammatory cascade, also have the potential to be useful.36–42 As of this writing, the United States Food and Drug Administration has issued advisories regarding celecoxib and valdecoxib, while rofecoxib and valdecoxib have already been withdrawn from the market by manufacturers, due to an increase of cardiovascular adverse events associated with long-term use, so caution is advised pending final regulatory declarations.

Minimizing Anesthesia-Controlled Time in the Operating Room

The potential value of the RA induction room should not be underestimated from the pharmacoeconomic perspective. Performing PNB techniques before OR entry has been shown to be associated with a time savings of approximately 9 min of OR time per case compared with using GAVA without blocks.9 In general, when the patient enters the OR ready for surgical preparation, and has a faster emergence (and exit) due to the use of sedation versus GAVA, an OR with five cases can save 45 min/day. If the cost of a minute of OR time is estimated to be $30, then 1 day of this amount of time savings for five cases carries a potential cost reduction of $1350. A portion of this theoretical savings is likely “real” in centers operating at or above 80% capacity, in which forced overtime (of preoperative, intraoperative, and postoperative nursing/ancillary staff) is a major budgetary expenditure. The cost savings becomes even more significant when cases later in the day (eg, after 3 pm or 5 pm) are commonly “stacked” into fewer available staffed ORs, further lengthening already-long clinical days. Dexter and colleagues estimated that emergence which is 6 min faster than baseline likely translates to a per-case overtime reduction ranging from 1.3 to 2.6 min.43 In a 50-case surgical pavilion, this translates to 65 to 130 min of overtime saved. How the reduction of anesthesia-controlled time influences OR staff retention is unknown. Repeated episodes of forced overtime for OR staff may lead to a loss of staff morale, which may translate to staff turnover. Staff turnover implies the replacement of departed, experienced staff with less experienced staff, associated training, and possible surgical process inefficiencies until training is complete and posttraining experience is sufficient to resume an optimally efficient surgical process. Prolonged anesthesia-controlled time may also adversely affect staff morale in both phase 1 and phase 2 PACUs. No studies to date have correlated reduced anesthesia-controlled time and perioperative nursing or OR staff retention.

OR time savings for outpatient shoulder surgery, via shorter induction time and emergence time values has also been substantiated when interscalene nerve block was used alone without GAVA.44 In this study, recovery room times and unplanned hospital admissions were also reduced in the patients treated with the interscalene block versus GAVA.44

Bypass of the Postanesthesia Care Unit in Ambulatory Surgery

In outpatient surgery, PACU bypass has been shown to be achievable in nearly 90% of patients receiving exclusively RA techniques (including neuraxial techniques if hemodynamic criteria are met).10 To achieve PACU bypass in an institution where RA techniques are used in high volume, it is important to use tested criteria that incorporate both physiologic parameters and immediate symptomatic outcomes to determine PACU bypass eligibility and to use criteria that incorporate specific outcomes for both RA patients and GA patients.45 Criteria that incorporate the original Regional Anesthesia PACU Bypass Criteria,10 the traditional Modified Aldrete Score,46 the White-Song Fast-Tracking Criteria,47 and the Mayo Modified Discharge Scoring System have been recently proposed (Table 74–1).45,48

In surgical pavilions with large caseloads (eg, 50 cases/day), an 80% PACU bypass rate (compared with no PACU bypass) can lead to a PACU nurse full-time equivalent (FTE) staffing reduction of up to four FTE if the PACU nurses are full-time employees, or by 20 nursing hours if the PACU nurses are part-time employees.43 When combined with forced overtime of OR staff and step-down recovery staff, OR time savings and PACU bypass (documented to be achievable with exclusive use of RA) can present important cost-saving opportunities for the hospital.

In a patient population undergoing anterior cruciate ligament (ACL) reconstruction, PACU bypass is associated with hospital cost reduction of $420.49 The cost savings component from the initial of $420 per PACU bypass patient (which excluded savings associated with nurse staffing reductions highlighted by Dexter and colleagues detailed earlier) were likely attributable to RA patients experiencing fewer symptoms than those receiving GAVA. Williams and coworkers also reported that throughout the multipavilion university hospital during peak use of PACU bypass (3000 outpatient orthopedics procedures per year), PACU nurse staffing requirements for 25,000 surgical patients per year (throughout all pavilions) consisted of 28 FTE PACU nurses.49 When the main campus multipavilion hospital relocated outpatient orthopedics to another off-site hospital in the health system, PACU bypass was used on the main campus for monitored anesthesia care cases only. Soon after, PACU nurse staffing requirements increased to 36 FTE for the same annual caseload.49 PACU bypass should not be underestimated as a potentially powerful cost management tool in ambulatory orthopedic anesthesia when nerve blocks are routinely used, not only from a staffing standpoint, but also with respect to overall symptom reduction and return to wakefulness during same-day recovery.

Successful Same-Day Discharge in Ambulatory Surgery

Woolhandler and Himmelstein have estimated that the cost of hospital admission (for all types of diagnoses and procedures) is $1050.50 Williams and coworkers found that the hospital cost increment associated with an overnight admission after ACL reconstruction was $385.49 The likely cost differences in the findings of Williams and coworkers versus those of Woolhandler and Himmelstein are likely related to the generally healthy status of outpatients presenting for ACL reconstruction. The key point is that it will always be less expensive for a patient to go home immediately after outpatient surgery than to be admitted overnight for observation.51 That said, precautions are required to ensure that costs are not incurred later in the form of requiring hospital readmission for complications improperly managed during the initial admission, especially since these readmissions are often ineligible for third-party reimbursement. Indeed, refractory pain is the most common cause of hospital readmission after discharge, accounting for over one third of such readmissions.52

It is important to understand that the associated hospital cost reductions of $420 for PACU bypass and $385 for successful same-day discharge were calculated by using standard econometric techniques of multivariate regression analysis.49 Thus, the associated cost savings captured in these values incorporate any and all expenditures related to OR time, additional RA equipment and medications used, prophylactic antiemetics, and postoperative parenteral nursing interventions required for symptoms. When specific time-resource and symptomatic outcomes were incorporated into the stated multivariate regression analysis, these covariates were not independent predictors of hospital costs, only PACU bypass and successful same-day discharge were independent predictors of hospital cost reductions. Thus, when deriving any cost analysis equation using these cost-saving values, it is important to use these values only for such analyses, pending the results of future, more detailed, economic studies. It would not be methodologically correct to simultaneously incorporate cost values from myriad other studies that calculate various itemized costs of events such as “minutes of PACU time,” “minutes of phase 2 recovery time,” or individual drug costs or labor costs. The use of cost estimate values from multiple studies is methodologically incorrect since “double-counting” would occur, which may artificially elevate incremental differences in cost, cost-benefit, cost-effectiveness, and cost-utility.

Successful Resource Management of Well-Trained Ra Practitioners in the Ambulatory Surgery Setting

In ambulatory surgery, well-trained RA practitioners may be tempted to implement a comprehensive nerve block care algorithm for all surgical patients. This enthusiasm should be tempered by the consideration of the “opportunity cost” of providing labor-intensive nerve block anesthesia for patients who may not necessarily benefit from these procedures. For instance, in the study by Williams and colleagues,11 543 patients underwent “relatively noninvasive” outpatient knee surgery, while the remainder underwent “more invasive” knee surgery. Forty-three percent (253/543) of the patients undergoing the noninvasive procedure received femoral (with or without sciatic) nerve blocks, but the use of nerve blocks in these patients was not associated with a reduction in symptoms, nursing interventions, or unplanned hospital admissions.11 As a result, these authors concluded that based on this retrospective review of a significant clinical caseload of noninvasive knee surgery outpatients, nerve blocks should be reserved for patients who have significant refractory postoperative pain, or in other special situations such as a complicated pain history or intolerance to traditional oral analgesic techniques.

In the same review of 1200 consecutive knee surgery outpatients, 657 underwent more invasive knee surgery, and 527 of these 657 (80%) received femoral with or without sciatic nerve blocks.11 In these patients, nerve block anesthesia and analgesia were significantly associated with reduced pain symptoms during recovery (therefore, fewer nursing interventions for pain management), and fewer unplanned hospital admissions. It is important to note that the selection of nerve block anesthesia–analgesia appears to be necessary but not sufficient to comprehensively reduce postoperative nursing interventions and unplanned hospital admissions: whenever GAVA was used (in the presence or the absence of nerve block anesthesia), the odds ratio of more associated symptoms after GAVA was 2.1 (p < 0.001), whereas the odds ratio of more associated unplanned hospital admissions after GAVA was 3.3 (p = 0.001).11 Thus, nerve block anesthesia and analgesia for indicated (more invasive) procedures (ideally in the setting of a comprehensive multimodal analgesic care plan) and the avoidance of GAVA (for all procedures) will likely provide the anesthesia care team (and the hospital) with the fewest possible side effects and the greatest facilitation of successful same-day discharge. However, routine nerve blocks for noninvasive knee surgery may be an investment of RA practitioners' time (and risk) that may provide relatively little benefit, and that takes away opportunity for RA practitioners to engage in other value-adding activities. Such value-adding activity may include the placement of continuous nerve block catheters for select patients, when time may have been only available previously to administer a single-injection nerve block.

Pain Risk Stratification

Upper Extremity Surgery

For outpatient shoulder or upper extremity surgery, there have been few substantiated, comprehensive recommendations for allocating nerve blocks (single-injection vs continuous infusion) based on anticipated postoperative pain, postoperative nursing interventions (with versus without), or unplanned hospital admissions.

In shoulder surgery, the comparison of single-injection blocks with GAVA has shown predictable findings. For arthroscopic acromioplasty of the shoulder performed under GAVA, Singelyn and colleagues showed that interscalene nerve block provided definitive recovery advantages over suprascapular nerve block, single-injection intraarticular local anesthetic, and controls.53 For outpatient open surgery of the rotator cuff, Hadzic and coworkers reported that GAVA use (vs single-injection brachial plexus block with ropivacaine) led to increased postanesthesia care unit admissions (vs phase 1 recovery bypass), higher reports of postoperative pain, longer time to ambulation, longer time to same-day discharge, and higher risk of unplanned hospital admission.54 In this study, no outcome differences occurred in follow-up from 24 h to 2 weeks after surgery, but this study was underpowered.

In a recent review, Boezaart suggested that an anterior approach to the brachial plexus is ideal for open-shoulder surgery, whereas a posterior (paravertebral) approach is well suited for arthroscopic surgery.55 In this review, Boezaart explains that the anterior approach to the brachial plexus concomitantly provides reliable anesthesia to the overlying skin, but the posterior approach does not. Potential advantages to the posterior approach may include less frequent blockade of the phrenic nerve, but the posterior approach is also associated with less motor block. Arthroscopic shoulder surgery using the posterior approach, most commonly, must be accompanied by general anesthesia,55 in order to provide sufficient analgesia to the overlying skin.

In two recent, separate studies of the efficacy of intraarticular analgesic infusions, some conclusions can be drawn regarding pain risk stratification in common outpatient shoulder procedures. One study by Harvey and associates showed that patients undergoing arthroscopic subacromial decompression of the shoulder receiving ropivacaine 0.2% in a continuous subacromial infusion experienced 34% lower pain scores than did saline controls.56 Another study by Boss and coworkers57 showed that patients receiving a continuous subacromial infusion of bupivacaine 0.25% after open acromioplasty and rotator cuff repair did not experience any difference in pain relief from those receiving saline placebo infusion. In one study by Klein and colleagues, patients undergoing open rotator cuff repair with an active treatment of local anesthetic infusion via an indwelling interscalene catheter had significantly improved pain management than did control patients receiving a saline placebo infusion.58 Thus, the logical conclusion can be reached that open acromioplasty and rotator cuff repair likely creates more postoperative pain than does arthroscopic subacromial decompression, based on the lack of responsiveness of the former (more invasive) procedure to subacromial infusion analgesia, whereas the less invasive procedure is responsive to subacromial infusion analgesia. However, open rotator cuff repair patients have favorable analgesic responses when a continuous interscalene catheter with local anesthetics is used.

A review by Chelly and associates59 provides an overview that may guide practitioners for categories of postoperative shoulder pain, until more definitive evidence is available. In this review, shoulder procedures are clustered into a catheter-eligible category if the following procedures are involved: shoulder arthroplasty, rotator cuff repair, Bankart repair, and open reduction/internal fixation of the humerus. The benefit of continuous interscalene catheters for shoulder arthroplasty and rotator cuff repair is well documented.58,60–65 Although logic would indicate similar effectiveness for less invasive procedures, there is little evidence at this time to indicate that interscalene brachial plexus catheters would be similarly useful for patients undergoing less invasive shoulder operations such as shoulder stabilization procedures, distal clavicle resection or acromioplasty, subacromial decompression, biceps tenodesis or tenotomy, or even routine debridement inside the glenohumeral joint, when compared with single-injection nerve blocks and perioperative multimodal oral analgesia. Thus, studies are needed to show the benefit of continuous nerve blocks (vs single-injection), and single-injection blocks (vs no blocks) for a wide variety of shoulder procedures that produce an uncertain magnitude of postoperative pain.

For outpatient wrist and hand surgery, Hadzic and coworkers addressed this patient population comparing chloroprocaine infraclavicular nerve block with GAVA, showing that GAVA use led to increased postanesthesia care unit admissions (vs phase 1 recovery bypass), higher reports of postoperative pain, longer time to ambulation, and longer time to same-day discharge.66 However, there were no outcome differences in follow-up from 24 h to 2 weeks after surgery. This latter finding was underpowered and did not show statistical equivalence.66 Chan and associates prospectively studied nonrandomized hand surgery patients (n = 126) undergoing either GAVA (n = 39), axillary block (n = 42), or Bier block (n = 45).67 GAVA was associated with the most postoperative symptoms and nursing labor intensity, as well as the longest discharge times. Bier block patients had the fastest recovery times and lowest associated total perioperative costs, but also were at small risk for conversion of the anesthesia plan (2/45) to GAVA due to tourniquet pain. Gebhard and colleagues retrospectively studied hand surgery patients (n = 62) receiving GAVA (n = 20), Bier block (n = 21) or wrist block (n = 21).68 They found that wrist block patients were discharged home soonest, and encountered (1) less hypertension than did Bier block patients; and (2) less hypotension than did GAVA patients.68 McCartney and coworkers prospectively studied 100 hand surgery outpatients randomized to receive GAVA (n = 50) or brachial plexus block with lidocaine (n = 50). These authors found essentially similar if not identical findings to those reported by Hadzic (prospectively) and Chan (retrospectively) as mentioned earlier. McCartney and coworkers also concluded that there were no long-term (2-week) pain outcome differences, although the brachial plexus block group only received the short-acting local anesthetic lidocaine.69

For other (distal) upper extremity surgery, Chelly and associates state that PNB catheters are likely indicated for implantation procedures after trauma, as well as for open reduction/internal fixation of the hand or digits,59 although a prospective randomized trial to definitively verify this intuitive concept may be difficult to achieve. Ilfeld and colleagues have shown that a continuous infraclavicular brachial plexus catheter (vs placebo catheter infusion) resulted in less postoperative dynamic pain and opioid consumption and fewer sleep disturbances.70 The surgical procedures performed included open reduction/internal fixation (elbow, radius, or ulna), bony/capsular wrist procedures (carpectomy, capsulodesis, fusion, or shrinkage) metacarpal arthroplasty, suspensionplasty, and ulnar nerve transposition. Although all of these procedures are intuitively painful, the small study sample size (30 patients divided equally between two groups) prevents the practitioner from distinguishing relative postoperative pain scores per procedure.70

Less invasive upper extremity procedures (typically applicable to outpatients) have not been comprehensively studied with respect to potential the value of continuous catheters versus single-injection blocks. However, Rawal and coworkers71 showed that an axillary continuous nerve block catheter with intermittent bolus dosing provided excellent wrist and hand analgesia for patients undergoing surgical procedures that may have been somewhat less invasive than those described earlier by Ilfeld and colleagues.70 Rawal and coworkers' study included 60 patients who received a mepivacaine axillary block bolus and concomitant nerve block catheter placement and were undergoing carpal tunnel release (n = 11), finger fracture repair (n = 11), tendon repair (n = 10), finger joint arthrodesis (n = 10), wrist arthroscopy (n = 8), and tumor resection or other procedures (n = 10). This study neither studied a control group of patients undergoing no nerve block, nor a treatment group undergoing single-injection nerve block only. In fact, 3 of the 60 patients studied did not use the bolus dose function postoperatively. However, most every bolus treatment was prompted by patients achieving a verbal pain score of at least 5 (out of 10), and bolus treatments returned pain scores to around 3 (out of 10), thus providing clinically significant analgesia.71 Patients in this study used all of their allotted boluses by the twelfth hour after surgery; as such, the reader may speculate that patients may simply benefit from a long-acting single-injection nerve block designed to provide 18 h of postoperative analgesia, instead of the technical complexity of a nerve block catheter. Rawal and coworkers found that the most common cause of patient dissatisfaction with the continuous catheter, intermittent–bolus technique was hand numbness.71 Therefore, patient uncertainty and potential dissatisfaction associated with prominent numbness or motor block should likely be factored into the decision of which nerve block technique is selected when a single-injection versus a continuous catheter technique is considered.

To summarize, outpatient hand surgery often results in significant postoperative pain, with patient pain scores often reaching or exceeding VAS 5 (out of 10).72 Retrospective reviews and prospective studies have demonstrated uniformly that patients receiving PNBs have significantly improved outcomes on the day of surgery compared with patients receiving GAVA. For shoulder surgery, the use of continuous brachial plexus catheters is sufficiently substantiated to recommend their routine placement (by trained practitioners) for invasive shoulder surgery. These findings also encourage the routine use (by trained practitioners) of all peripheral/regional techniques for upper extremity surgery, although further research is needed to determine outcome benefits of continuous versus single-injection nerve blocks in the days and weeks following mildly to moderately invasive shoulder and distal upper extremity surgery. The usefulness of intraarticular and incisional infusions after simple arthroscopic procedures of the shoulder has been documented, but these articular and incisional infusions do not appear to confer sufficient analgesia after more invasive, open, shoulder surgery.

Knee Surgery

Recommendations for rational nerve block selection (single-injection, vs continuous catheter, vs none) in outpatient knee surgery have been recently suggested.73 This guideline incorporates the resource management principles described earlier, and creates three major categories: (1) noninvasive, (2) more invasive, and (3) most invasive (Table 74–2). Noninvasive implies that routine use of nerve block analgesia is probably not necessary, as described previously.11 “More invasive” implies that a routine femoral nerve block would be recommended, but that a sciatic nerve block is probably not necessary, since the vast majority of the postoperative pain is likely attributable to the femoral nerve distribution. “Most invasive” implies that the postoperative pain will be likely attributable to both femoral and sciatic nerve distributions, and both nerves would likely benefit from routine blockade. The more invasive and most invasive knee surgery categories are based on clustering moderate and severe into one category of surgical invasiveness.11 This algorithm73 also describes when a single-injection block would likely be sufficient, vs when a continuous nerve block catheter would likely be of greater benefit than a single injection.

Foot and Ankle Surgery

For outpatient foot and ankle surgery, few if any substantiated, comprehensive recommendations are available for allocating nerve blocks (single-injection vs continuous infusion) based on anticipated postoperative pain, postoperative nursing interventions (with vs without), or unplanned hospital admissions. Interestingly, few studies compared PNBs with GAVA for surgery below the knee. In addition, few studies have categorized postoperative foot and ankle pain as sufficiently manageable with a single-injection nerve block vs requiring an indwelling sciatic nerve catheter. Most studies evaluating the use of various approaches to the sciatic nerve block for foot and ankle surgery have either simply evaluated block success rate (with no comparative treatment group), compared varying popliteal/sciatic nerve block approaches (sometimes with neuraxial techniques), or compared continuous infusion strategies.

Singelyn and colleagues provided one of the first studies describing efficacy of the popliteal fossa block placed with nerve stimulator guidance, and reported a low 3% (15/507) rate of conversion to GAVA,74 whereas Provenzano and coworkers reported a conversion rate to GAVA of 18% (84/467).75 In the same study, Provenzano and coworkers reported a significant reduction in postoperative opioid requirements in patients with a successful popliteal fossa block compared with 367 patients who did not receive the block.75

When neuraxial techniques are being considered, Curatolo and associates provide important insight that epidural anesthesia for foot and ankle procedures is associated with a high (4.4%, 7/160) conversion rate to GAVA, although epidural success was correlated with larger per-segment doses.76 Two studies have compared the efficacy of spinal anesthesia with popliteal fossa block, and both showed intraoperative efficacy of both techniques, although the in-hospital recovery after spinal anesthesia was more lengthy,77,78 with an additional risk of urinary retention after spinal.78 Generally speaking, meaningful postoperative analgesia should not be expected after neuraxial techniques, and analgesia after popliteal fossa block will depend on the local anesthetic agent (and additives) used. In addition, when the analgesic duration of the popliteal sciatic block (with long-acting local anesthetics) is compared with the durations of ankle block, foot block, or subcutaneous infiltration, one should expect doubling or tripling the analgesic duration with a sciatic-specific depot injection.79

Continuous sciatic nerve catheters have gained popularity in recent years. After being introduced by Singelyn and colleagues who described a complex Seldinger (catheter-over-guidewire) technique for catheter placement (achieving a 92% success rate),80 authors have repeatedly found that continuous infusion nerve blocks lead to excellent analgesic outcomes when compared with single-injection blocks (or placebo catheters).81–83 At this time, there are no definitive guidelines for the selection of single-injection vs continuous popliteal/sciatic techniques based on anticipated surgery, with the exception of recommendations implied by the myriad findings reported earlier. Others have suggested that for hardware removal from the foot and ankle, a single-injection block is sufficient, whereas for most other foot ankle procedures, the use of a continuous sciatic catheter may have additional benefits.59 The ultimate decision about the use of single-injection vs continuous sciatic techniques may depend on realistic expectations by the surgeon regarding return to weight-bearing status. In this situation, the surgeon may not have an accurate impression of weight-bearing success rates in the setting of uncontrolled pain, but full weight-bearing in a the setting of a partially anesthetized sciatic nerve (via a continuous sciatic catheter) may be ill-advised.

Summary Statement Regarding Nerve Block Technique Allocation

The rational use of single vs continuous nerve block techniques in ambulatory surgery, that is, allocating the “scarce resource” of the well-trained RA practitioner and avoiding routine nerve blocks for patients who probably do not need them (or providing single-injection nerve blocks when a continuous catheter is not likely needed) will free up time to perform catheter techniques for patients undergoing indicated procedures. In OR settings that care for both outpatients and inpatients, this strategy allows the skilled RA practitioner to also dedicate time to nerve block catheter placement where it is most beneficial.

The Need for Training & the Importance of Multimodal Analgesia

Despite the evolution of multimodal analgesia and nerve block analgesia, little progress has been made with respect to patient perceptions of quality improvements in analgesia care. This has been documented in two studies almost a decade apart using similar methodologies by Warfield and Kahn84 and Apfelbaum and coworkers.85 Warfield and Kahn surveyed hospital patients from 300 hospitals (42% of which had acute pain management programs), and found that 77% of patients experienced pain after surgery, with 80% of these respondents categorizing their pain as moderate to severe. In the study by Apfelbaum and coworkers 8 years later, 80% of surveyed adults (n = 250) experienced pain after surgery, with 86% of these respondents characterizing the pain as moderate, severe, or extreme.85 This latter study did not characterize the hospital-based acute pain management infrastructure where respondents underwent surgery.85 Thus, the time between the first development of federal recommendations for acute pain management in 199286 and pain management mandates by the Joint Commission for Accreditation of Health Care Organizations in 200187 did not show perceptions of improvement in pain management after surgery in the general population.

One approach to making significant progress toward improving postoperative pain management is to properly educate clinical staff. One study by Loder and colleagues describe the surveying of the clinical staff of a rehabilitation hospital about their knowledge and attitudes regarding effective pain management.88 This study showed that rehabilitation hospital staff hold generally progressive attitudes toward the treatment of pain, but with a substantial degree of ambivalence about the use of opioids in that treatment.88 The same staff rated their own lack of education about pain management as one of the chief barriers to effective pain management, and a large percentage reported feeling uncomfortable with various technical aspects of pain care.88 As pain management techniques continue to increase in both effectiveness and complexity, detailed planning and education is required for all health care providers, patients, and families (surgical hospital, rehabilitation hospital, outpatient rehabilitation facilities, and outpatients at home) to help make meaningful analgesia a successful endeavor.

Opioid protocols have been successfully implemented in rehabilitation hospital settings. Cheville and associates reported on the successful implementation of the use of controlled-release oxycodone for rehabilitation inpatients after unilateral total knee replacement surgery in a randomized controlled trial.89 In this study (n = 59), patients were randomized to receive controlled-release oxycodone vs placebo every 12 h, with the opportunity to receive immediate-release oxycodone every 4 h as needed for breakthrough pain. Patients in the treatment group had lower pain scores, significantly better range of motion of the knee, and improved quadriceps strength than did the placebo group.89 In addition, treatment group patients were discharged home from the inpatient rehabilitation facility 2.3 days sooner than placebo patients.89 To summarize, a straightforward treatment intervention of a regularly scheduled controlled-release oral opioid led to significant outcome improvements and lower use of health care resources compared with the likely typical care plan of analgesics only as needed. However, such an approach may not gain widespread acceptance due to (likely unfounded) fears of opioid dependence and diversion. In addition, escalating doses of oral opioids are well associated with increasing side effects. Dose-dependent opioid-related side effects were well described recently in a study showing that the increase of daily opioid dosing by the equivalent of 4 mg of morphine is associated with one additional clinically meaningful opioid-related symptom, or one additional patient-day with an opioid-related clinically meaningful event.90 These symptoms and events specifically included nausea, vomiting, constipation, urinary difficulty, difficulty with concentration, drowsiness, light-headedness, confusion, fatigue or weakness, pruritus, dry mouth, and headache.90

Logic would dictate that using multimodal analgesia, with a PNB technique as an integral component, will very likely help improve patient perceptions of pain relief after surgery, especially when the patients are no longer in a surgical hospital (eg, inpatient rehabilitation or discharged to home). Kehlet has reviewed important principles of multimodal analgesia on postsurgical physiology.91 However, neuraxial regional anesthetic techniques are unlikely to be a meaningful contributor to postoperative analgesia out of the surgical hospital setting, for obvious reasons. PNB anesthesia, including continuous perineural infusions, will logically serve as a future centerpiece of multimodal analgesia outside the surgical hospital setting in the years to come, but significant progress will need to be made with respect to infusion technology; dose-response curve derivation; and education of clinical staff, patients, and family members when an anesthesiologist is not in immediate attendance.

Decreasing Role for Neuraxial Techniques

For many years, neuraxial anesthesia was highly recommended as a technique of first choice for patients undergoing lower extremity total joint replacement or hip fracture repair. Much of this enthusiasm was based on the classic meta-analysis by Sorenson and Pace, which reported that the use of neuraxial vs general anesthesia was associated with a significant reduction in mortality (odds ratio = 0.67), and a 31% reduction in deep venous thrombosis, for the repair of fractures of the femoral head.92 Years later, the meta-analysis of several types of surgery (including orthopedics) by Rodgers and coworkers reported significant reductions in mortality (odds ratio = 0.70), pulmonary embolism (odds ratio = 0.45), deep vein thrombosis (odds ratio = 0.56), myocardial infarction (odds ratio = 0.67), renal failure (odds ratio = 0.57), pneumonia (odds ratio = 0.61), respiratory depression (odds ratio = 0.41), and transfusion requirements (odds ratio = 0.45 to 0.50) when neuraxial anesthesia was used instead of GAVA with postoperative intravenous patient-controlled opioid analgesia.93

Regional Anesthesia Techniques in the Setting of Modern Anticoagulants

Although the aforementioned studies would logically serve as landmarks for change in clinical practice, enthusiasm for neuraxial techniques has significantly decreased with the introduction of more potent and effective anticoagulants in routine clinical practice. Guidelines for neuraxial anesthetic use in the setting of systemic anticoagulation have been well publicized,94,95 and the risks have been more accurately defined.96 However, it seems less likely that patients will have neuraxial anesthesia available to them in these clinical settings in the near future, pending a long-term outcome study that describes the frequency of adverse events created (eg, hemorrhagic cerebral infarction97) versus adverse events prevented (eg, pulmonary embolism and death) by the use of newer antithrombotics and anticoagulant agents.

In a meta-analysis of outcomes after the use of fondaparinux (the first agent in the new class of factor Xa inhibitors) versus enoxaparin,98 fondaparinux was shown to reduce the odds of venous thromboembolism by 55% compared with enoxaparin. Although fondaparinux was associated with higher “major bleeding,” this did not translate to a clinically adverse outcome such as risk of reoperation or death. In addition, the use of fondaparinux appeared to offset any outcome benefits achieved by neuraxial anesthesia vs GAVA, specifically for the avoidance of venous thromboembolism. The likely loss of the availability of neuraxial techniques as surgeons focus on the prevention of venous thromboembolism, without regard for the other documented benefits of neuraxial anesthesia/analgesia (eg, on the cardiac, pulmonary, and renal organ systems), forces the well-trained RA practitioner to incorporate continuous PNBs to give patients any chance for meaningful postoperative analgesia after lower extremity joint replacement or fracture repair.

Evolving Role for CPNB in Lower Extremity Surgery

The use of continuous PNBs for joint replacement has been studied in recent years with uniformly positive results. In the early 1990s, outcomes were more equivocal, but many studies during that period addressed primarily single-injection techniques. In one study by Hirst and colleagues, patients receiving general anesthesia did not gain additional analgesic benefit from a femoral perineural infusion compared with patients who received a single-injection nerve block, although both groups required less postoperative opioid analgesia than did controls who received no block.99 A study by J. G. Allen and associates showed that patients receiving femoral-sciatic single-injection blocks had better short-term (24 h) pain outcomes than did patients only receiving spinal.100 Another single-injection block study for total knee replacement patients by H. W. Allen and colleagues showed that the addition of single-injection sciatic nerve block did not confer any additional benefit vs the use of femoral nerve block alone.101 This latter finding has not been reproduced in any study with similar methodology.

In the late 1990s, two important studies from Europe examined rehabilitation outcomes after total knee replacement when a continuous femoral catheter was used (vs epidural catheter or control IVPCA device). These studies were predicated on the notion that continuous femoral nerve block analgesia (vs IVPCA) leads to not only better pain relief but also significantly better knee flexion, faster achievement of ambulation goals, and overall faster convalescence. Studies by Capdevila and coworkers102 and Singelyn and associates103 showed that total knee replacement patients undergoing general anesthesia with either continuous epidural analgesia or continuous femoral nerve block analgesia made faster progress meeting rehabilitation objectives and were discharged from the inpatient rehabilitation unit sooner than were patients receiving IVPCA. Patients receiving femoral nerve catheter infusions experienced fewer side effects than did epidural patients in both studies, and continuous femoral catheter patients were discharged home from inpatient rehabilitation units sooner by 20% (40 vs 50 days;102 and 17 vs 21 days103 in the femoral catheter vs IVPCA groups).

In the United States, a similar anesthetic treatment method was applied by Chelly and colleagues to total knee replacement patients.104 All patients received GAVA and were randomized to receive IVPCA, epidural infusion, or single-injection femoral-sciatic blocks followed by a continuous femoral infusion. Patients with continuous femoral blocks (vs IVPCA patients) had an associated reduction of postoperative bleeding by 72% (p < 0.05), achieved better performance on continuous passive motion, had a 90% decrease in serious complications (including less blood loss), ambulated sooner (2.5 vs 3.5 days), and had a 20% decrease in the length of hospitalization (4 vs 5.5 days).104 Duration of hospitalization did not include postoperative long-term rehabilitation (which is usually done on an outpatient basis in the United States)104 as did the two previously listed European studies,102,103 although early postoperative rehabilitation was aggressive.

The question of the need for a continuous sciatic catheter after total knee replacement was recently addressed in a prospective pilot study by Ben-David and coworkers.105 In this study, 12 consecutive patients had continuous femoral and sciatic catheters placed preoperatively, but only the femoral catheters were dosed to allow for postoperative evaluation of sciatic nerve function (specifically aiming to produce dorsiflexion). Sciatic catheters were dosed only if dorsiflexion was intact and postoperative knee pain was refractory to additional boluses through the femoral nerve catheter. Ten of these 12 patients required dosing of the sciatic catheters, with median pain scores of 7.5 (out of 10) before dosing, and 2.0 after dosing. This pilot study refutes the earlier finding by H. W. Allen and colleagues101 with respect to the value of sciatic nerve block for total knee replacement.

For hip surgery, there has been less convincing evidence with respect to PNBs, pain management, and rehabilitation outcomes. From the Cochrane Database of Systematic Reviews (2002, with no updates since then), Parker and associates report that “Because of the small number of patients included in this review and the differing type of nerve blocks and timing of insertion, it is not possible to determine if nerve blocks confer any significant benefit when compared with other analgesic methods as part of the treatment of a hip fracture. Further trials with larger numbers of patients and full reporting of clinical outcomes would be justified.”106 However, this systematic review did not appear to include studies reported later in this section.

Fournier and coworkers reported that patients receiving single-injection femoral nerve block vs sham block had a 4-h delay in requests for first parenteral analgesic after prosthetic hip surgery, although pain outcomes were no different at 24 and 48 h.107 Stevens and colleagues reported that patients undergoing total hip arthroplasty receiving lumbar plexus single-injection blocks had less pain for up to 6 h after surgery and less blood loss during and for up to 48 h after surgery.108 Naja and associates reported (retrospectively) that elderly hip fracture patients receiving lumbar plexus and parasacral blocks (vs GAVA) encountered significantly less hypotension during surgery, were less likely to be admitted to the intensive care unit after surgery (0/30 vs 11/30), and had a shorter length of hospital stay (7 vs 14 days).109 De Visme and colleagues (n = 15)110 and Buckenmaier and coworkers (n = 10)111 provided the first two reports of lumbar plexus and parasacral plexus blocks as the sole anesthetic for hip surgery (ie, coadministered GAVA or spinal anesthesia were not deemed necessary). Souron and associates found that patients receiving intrathecal morphine (0.1 mg) vs single-injection lumbar plexus block with ropivacaine had better overall analgesic outcomes with intrathecal morphine after primary hip arthroplasty, although intrathecal morphine patients had a 37% incidence of urinary retention (vs 11% for lumbar plexus block (p < 0.05).112 Biboulet and colleagues compared single-injection femoral and lumbar plexus blocks with intravenous opioid PCA for patients undergoing total hip arthroplasty, and found that PCA was as efficient as the single-injection blocks used in overall outcome, even though the pain scores and opioid requirements were lowest in the lumbar plexus single-injection group during the first 4 h after surgery.113 Kullenberg and coworkers prospectively studied 80 patients who underwent hip fracture repair and were randomized to receive single-injection femoral nerve block (n = 40, block placed postoperatively) or no block (n = 40).114 This study found that nerve block patients had 15 h of meaningful pain relief and ambulated 13 h sooner (23 vs 36 h) than did patients who received no block.114

All of the above-mentioned studies particularly addressed single-injection nerve blocks for hip surgery and did not address continuous nerve block catheters. The following two studies did specifically address perineural infusions for such patients. Singelyn and Gouverneur prospectively evaluated 1338 patients undergoing total hip arthroplasty who chose either IVPCA, continuous femoral block, or patient-controlled epidural analgesia.115 These authors reported that patients who received continuous femoral infusion had the highest satisfaction, fewest side effects, the lowest request rate for supplemental opioids, and the fewest technical problems, but pain scores themselves did not significantly differ.115 In a randomized clinical trial Turker and coworkers compared patients receiving continuous lumbar plexus catheter (n = 15) with others receiving epidural catheter (n = 15) for hip hemiarthroplasty under GAVA and found that lumbar plexus catheter patients (1) had less motor block, (2) ambulated sooner, and (3) had significantly fewer overall complications.116

Based on the available evidence, femoral or lumbar plexus single-injection blocks have the potential to improve immediate postoperative analgesic outcome and possibly reduce hospital resource utilization in patients having invasive hip surgery. Unlike single-injection femoral nerve blocks, single-injection lumbar plexus blocks are commonly associated with a risk of epidural spread, which carries a risk of perioperative implications in a frail elderly patient with limited cardiovascular reserve. One theoretical strategy to avoid precipitous hypotension (from epidural spread with a single-injection lumbar plexus block) is to provide slow incremental doses via a continuous lumbar plexus catheter, helping to potentially prevent epidural spread. In this setting, the incremental doses can be stopped as soon as it appears that surgical anesthesia/analgesia seems to be in place, but still keeping the catheter injection port available to the anesthesia team during surgery in the event additional boluses are needed intraoperatively.

Other than epidural spread, the other potential concern for lumbar plexus blocks (specifically continuous catheters) is the risk of hematoma in an anticoagulated patient. Technically more challenging continuous lumbar plexus blocks have been associated with hematoma-related complications in three patients documented in two case reports.117,118 In one patient who was fully anticoagulated and receiving aspirin (325 mg daily), three serial single-injection lumbar plexus blocks were given, the last of which was considered “difficult,” and a retroperitoneal hematoma was diagnosed a few days later.118 A second patient developed retroperitoneal hematoma in which significant vascular trauma was noted at the time the block was performed and the catheter placed.117 A third patient developed retroperitoneal hematoma 2 days after the single-injection psoas compartment block was placed, and this patient was rendered supratherapeutic on intravenous heparin throughout the 2 days after surgery.117 In each of these three cases, the clinical situations could not be reasonably considered as routine, and all of these cases resolved without sequelae and without requiring surgical evacuation. Other than these three cases, there have been no reports of hemorrhagic complications in patients receiving lumbar plexus catheters in the setting of modern anticoagulation practice.119 There have been no published data regarding the estimated risk of retroperitoneal hematoma associated with lumbar plexus catheter removal in patients receiving effective postoperative anticoagulant therapy for the prevention of deep venous thrombosis. More cases are required to establish the safety of such a protocol. When considering the risk of bleeding associated with the removal of a lumbar plexus catheter in patients who are anticoagulated for the prevention of DVT, it is also important to acknowledge that the consequences of such bleeding are less serious than the bleeding that may result from the placement or removal of an epidural catheter. Epidural hematoma has been associated with serious neurologic complications in over 50% of cases, whereas no serious neurologic complications have been associated with the development of a retroperitoneal hematoma. Nevertheless, the consensus of most clinicians who often utilize lumbar plexus and other continuous nerve block technique seems to be that removal of continuous nerve block catheters at this time need not receive the detailed scrutiny that has been deemed appropriate for epidural catheters.

Role of CNB Catheters & Effects on Rehabilitation in Upper Extremity Surgery

Data on the evaluation of physical therapy outcomes after major shoulder surgery performed with nerve block anesthesia are very limited. Two recent studies addressed in more detail the recovery of motor function, comparing different CNB infusion drugs and concentrations. Borgeat and colleagues studied hand motor function and the presence of paresthesias in the fingers after open-shoulder surgery, when pain was managed via a brachial plexus CNB. Two infusion drugs were compared: bupivacaine 0.15% and ropivacaine 0.2%.62 Infusions were run via an electronic device at 5 mL/h, with a 4-mL bolus available with a 20-min lockout. Although both treatment groups' patients had equal analgesia and were equally satisfied, bupivacaine CNB patients had more hand weakness at 24, 48, and 54 h while reporting more finger paresthesias up to 48 h after the initial nerve block than did the ropivacaine patients.62

Another recent study addressing motor function was conducted by Casati and coworkers.120 Patients who underwent open-shoulder surgery were assigned to one of two CNB infusion drugs: ropivacaine 0.2%, or levobupivacaine 0.125%. The infusion was run via an electronic device at 6 mL/h and allowed for a 2-mL bolus with a lockout of 15 min. There were no clinical differences in pain relief quality or motor function during the 24-h infusion period. Thus, it appears that ropivacaine 0.2% and bupivacaine/levobupivacaine 0.125% are equipotent in providing CNB analgesia in the brachial plexus. The infusion rate of 5 to 6 mL/h while allowing for intermittent boluses of 2 to 4 mL with a 15- to 20-min lockout provides an optimal balance of analgesia and preserved motor function.

One important safety feature to consider in the postoperative recovery of patients receiving continuous interscalene analgesia is adequate ventilatory function. It is generally accepted that most patients undergoing brachial plexus nerve blocks for shoulder surgery will encounter simultaneous block of the phrenic nerve, which is responsible for proper function of the diaphragm.121–124 Borgeat and associates measured respiratory function during use of a CNB dosing technique that used both a continuous infusion (ropivacaine 0.2%, 5 mL/h) and a CNB bolus function (3–4 mL with a 20-min lockout).125 In this study, all patients received a preoperative bolus injection of ropivacaine 0.75%, 30 mL, and all patients underwent major shoulder surgery (rotator cuff repair, n = 26; arthroplasty, n = 7). The control group consisted of patients receiving IVPCA with opioids. Patients in the CNB group had better pain relief for up to 24 h after surgery than the patients who were randomized to the opioid PCA. An important new finding was that overall respiratory function was better in the CNB group than in the PCA group. Forced respiration (ie, movement in the diaphragm on the nonoperated side) was better in the CNB group at 24 and 48 h than in the PCA group. The rationale for this finding was that the pain control was better in the CNB group, and that there were fewer opioid-related side effects (eg, respiratory depression) in the CNB group, facilitating patients' forced respiratory efforts. Interestingly, forced diaphragmatic excursion on the side of surgery was not significantly different between the CNB and PCA groups at 24 and 48 h after surgery. This study showed that forced respiratory effort was improved up to 48 h after surgery in the CNB group, which when combined with better analgesia in the CNB group, provided an important safety validation in the evolution of same-day discharge of patients with CNB catheters and appropriate infusion devices after shoulder surgery.

To date, there do not appear to be any reviews regarding the role of no block, versus single-injection or continuous nerve blocks of the brachial plexus, as it relates to the timing of meeting physical therapy objectives. Generally speaking, rotator cuff repairs are associated with a very deliberate and conservative physical therapy course, in which full range of motion is not attempted until 6 weeks after surgery, allowing for the surgically augmented insertions to heal. Meanwhile, therapy objectives after total shoulder arthroplasty or hemiarthroplasty are somewhat less conservative, in which full range of motion is attempted at 4 weeks after surgery. For both rotator cuff repair and shoulder arthroplasty, heavy lifting is usually not permitted until 3 months after surgery. As a result, hand movement is the only available proxy indicator of physical therapy progress, and this entails typically only the first few days after surgery while an indwelling interscalene catheter is infusing. A similar absence of data also appears to be the case as it relates to physical therapy function of the elbow, wrist, and hand.

Cost Analysis Illustrations Based on Available Data

Cost-Minimization Analysis

For this analysis, we compare three anesthesia treatments for a traumatic, closed patellar fracture in one patient. The mathematical formulas described here can be applied to actual incidences of compared outcomes using weighted-average techniques derived from larger surgical population data.126 For the individual patient in this example, GAVA is compared with two spinal anesthesia techniques, one of which uses a single-injection femoral nerve block (two-needle technique), the other which uses a femoral perineural catheter with single-injection sciatic nerve block (three-needle technique). Assumptions are as follows: (1) the GAVA patient will encounter a routine recovery room stay, then hospital admission for 2 days (pain management, antiemesis, and resolution of somnolence); (2) the two-needle patient will encounter an unplanned hospital admission to “23-h observation” after experiencing posterior knee pain treated with opioids and leading to PONV; and (3) the three-needle patient will be discharged home the same day with a disposable continuous infusion device after successfully bypassing the PACU. Based on cost assumptions (Table 74–3) described in previous reports,8,49 the sole task of cost-minimization shows the incremental cost savings is roughly $800 for using the two-needle technique, and an additional $200 when the added cost of equipment and medication for the three-needle technique is offset by the cost savings of PACU bypass and the avoided hospital admission.

Cost-Effectiveness Analysis

For a hypothetical CEA analysis from the patient's (and the societal) perspective, we shall assume that patients are most interested in returning to nonstrenuous work as soon as possible, with sufficient cognitive function and pain control. In what should be interpreted as a strictly speculative illustration, the GAVA patient is assumed to require 10 days to return to work due to lingering effects of volatile agents, cumulative dose of opioids, and sedative effects of antiemetics. The two-needle patient would probably have less cognitive dysfunction from (1) avoiding volatile agents and (2) encountering less cumulative effects of opioids and antiemetics, whereas the three-needle patient is pain-free with the clearest cognition and is able to return to work on postoperative day 3 (with an infusing perineural catheter). The increments of improvement in the return-to-work parameter are calculated by determining the difference in days to return-to-work from the reference value of 10 days (for the GAVA patient). Then the incremental differences are inserted in the denominator, and the incremental hospital costs comprise the numerator. The ratio of incremental hospital cost savings to day of work saved for the patient (see Table 74–3) was most cost-effective (ie, the lowest cost effectiveness ratio) when the decision was made to use a femoral catheter and sciatic single-injection block (three-needle technique), instead of just a femoral block (two-needle technique). It should be understood that an isolated cost-effectiveness ratio is only meaningful in the context of comparative methods being evaluated. In addition, hospital costs for each anesthetic technique do not clearly reflect the costs associated with anesthesiologist labor because the incremental workload is increased from GAVA, to the two-needle technique, to the three-needle technique, respectively. One would hope that well-trained RA practitioners recognize the value of their services and compassionately care for patients in such a way that gives the patient the best chance for immediate return to work in a health care system that is otherwise optimized.

With respect to using more population-based return-to-work data for a given procedure in an institution, certainly, gaussian curves illustrating the 95% confidence intervals of actual return-to-work outcomes could be incorporated into the calculation, and similarly 95% confidence intervals of the cost-effectiveness ratio can be determined to evaluate the extent of overlap between techniques for the desired parameter of effectiveness. However, population-based return-to-work data after the wide variety of orthopedic procedures performed are scarce, and available population data are likely based currently only on a vast preponderance of cases being performed under GAVA without any PNB techniques.

Cost-Benefit Analysis

Using the design for the CEA example discussed earlier, the effectiveness variable for CBA analysis now takes on a monetary value in the denominator. In this analysis we will assume the patient requires reasonable ambulation capabilities at work as a factory foreman paid an hourly wage, who is not paid for time off for medical leave. The return to work benefit outcome takes the form of achieved wages. The return to work timeline remains the same as that described in the earlier CEA example (see Table 74–3). In this example, the reference value for 10 days of lost wages is $2000 (after receiving GAVA with no blocks). Patients receiving spinal anesthesia with femoral nerve block are assumed to return to work on postoperative day 8, incrementally reclaiming $400 of the lost $2000 in wages. Patients receiving spinal, femoral nerve block catheter, and single-injection sciatic block are assumed to return to work on postoperative day 3, thus reclaiming $1600 of the lost $2000 in wages. These reclaimed wages are inserted into the denominator of the incremental cost-benefit equations, with the respective hospital cost increments inserted into the numerator. The cost-benefit ratios (CBR, see Table 74–3) indicate that the selection of the three-needle technique provides a better (lower) CBR than does the two-needle technique, when the reference GAVA technique is considered as the standard for comparison.

CUA involves the insertion of a patient-reported effectiveness variable into the denominator of the cost analysis. We will use the QoR-40 of Myles and colleagues127 as a measure of utility from the patient's perspective, again comparing the hospital costs of the GAVA patient ($1250 cost), vs $460 and $290 for spinal anesthesia patients receiving single-injection femoral nerve block vs continuous femoral and single-injection sciatic nerve blocks, respectively. The achieved utility scores based on the QoR-40 by Myles and colleagues (with possible highest score of 200) are assumed to be 170 (GAVA), 175 (two-needle), and 180 (three-needle), respectively. These QoR-40 scores are based on our unpublished data for 270 patients undergoing anterior cruciate ligament reconstruction with no nerve block, vs single-injection femoral block, vs continuous femoral infusion. The incremental cost-utility ratios would be described as “hospital costs per preserved QoR-40 score unit (see Table 74–3). The two-needle technique shows an incremental cost-utility ratio of 158:1 versus GAVA, and the three-needle technique cost-utility ratio is 96:1 vs GAVA.

In both the cost-benefit and cost-utility illustrations, as with the cost-effectiveness illustration, the decision to use the three-needle technique instead of the two-needle technique provided the most impressive incremental pharmacoeconomic benefit. For the cost-effectiveness and cost-benefit analyses, the analyses showed that the selection of the three-needle technique vs the two-needle technique provided a magnitude of “patient/societal benefit” by a factor of 14 (in comparison with the selection of the two-needle technique vs GAVA), whereas the cost-utility benefit, the magnitude of benefit was a factor of almost 5 when comparing the same increments per technique. Indeed, it is conceivable that a win-win anesthetic technique (eg, total RA incorporating nerve block catheters) may provide much greater societal benefit than the patient may be able to recognize if using only symptom-specific patient satisfaction surveys.

Obviously, the preceding examples are quite simplistic and do not account for the significant variability in the care of many patients. However, the core structure of the analyses, including multiple variables that are important contributors to costs and outcomes, can be analyzed using decision analysis trees and weighted-average techniques, to provide meaningful comparisons of anesthesia care techniques.

In the pharmacoeconomic analysis of the value of RA, several considerations are important. The first is to distinguish the objectives of outpatient surgery from the objectives for inpatient care. The second is to recognize that almost every decision made by the anesthesiologist may have a crucial influence on patient outcome not only in the short term, but also potentially in the long term. The third is for the skilled RA practitioner to rationally allocate his or her own labor-related resources to ensure that patient-specific and procedure-specific criteria are met for each and every patient being considered for regional anesthetic techniques, in order to minimize wasted effort on patients unlikely to benefit from these labor-intensive interventions. The fourth is that the anesthesiologist is likely the primary patient advocate with respect to analgesic outcome, due to less likelihood of expertise or even interest among surgical colleagues and other health care personnel. The fifth is that a coordinated effort is required to gain the surgeon's and patient's confidence in the anesthesia team while simultaneously guiding all other related health care personnel with education efforts to redefine policies and procedures. Finally, awareness of overall cost implications, to both the hospital and to society at large, is required to justify the expansion of (and reimbursement for) RA services to hospital administrators (and third-party payers or legislative entities).