Purpose
Stenosing tenosynovitis (STS) is a common condition treated by hand surgeons. Limited evidence exists to support the nonsurgical management of STS. The purpose of this study was to prospectively evaluate a cohort of patients with STS, and to determine the strategy for treating patients with this condition that is most cost effective in terms of dollars reimbursed by payers.
Methods
Prospective data were collected on patients diagnosed with STS between March 2014 and September 2014. All patients were initially treated with a corticosteroid injection. Patients with persistent symptoms were given the option of injection or surgery. A maximum of 3 injections were offered. All patients were evaluated every 6 months through office appointments or phone calls. A cost analysis was performed in our cohort using actual reimbursement rates for injections, initial and established patient visits, and facility and physician fees for surgery, using the reimbursement rates from the 6 payers covering this patient cohort. Cost savings were calculated based on offering 1, 2, and 3 injections.
Results
Eighty-eight digits in 82 patients were followed for an average of 21.9 months (range, 18.7–22.7 mo) after an initial corticosteroid injection. Thirty-five digits went on to surgical release, whereas 53 digits were treated nonsurgically. Had all patients initially undergone surgery, the cost would have totaled $169,088.98 ($1,921 per digit). Offering up to 3 injections yielded a potential savings of $72,730 ($826 per digit) or 43% of the total cost. For the 33 patients who underwent more than 1 injection, offering a second injection yielded potential savings of $15,956 ($484 per digit, 22.7%), and for the 7 patients presenting a third time, a third injection saved $1,986 ($283 per digit, 14.5%).
Conclusions
Based on the data from our cohort, the efficient way to treat STS in terms of health care dollars spent is to offer up to 3 injections before surgical release. The first injection had the highest component of cost savings, at $826 per digit.
Type of study/level of evidence
Economic/Decision Analysis III.
Key words
With the evolution of the American health care system, there has been a recent emphasis in the literature on practicing cost-conscious medicine.
1
, 2
, 3
Stenosing tenosynovitis (STS) is one of the most commonly treated conditions by hand surgeons, but no clear evidence exists to guide the treatment of this condition in terms of minimizing health care spending. A recent national study demonstrated that 74% of patients presenting with STS are initially treated with an injection, but that the majority of treating physicians attempted only 1 injection before surgery.4
Although there is some support to only giving a single injection of corticosteroid before surgical treatment of STS,5
giving up to 3 injections of corticosteroid for the treatment of STS is safe and has also been shown to be effective in preventing progression to surgery.6
Complications of injections are mild and generally self-limited, with no known reported cases of flexor tendon rupture related to injections.7
Given the low morbidity of corticosteroid injections and its relatively low cost, the tendency of surgeons to offer only a single injection before surgery may lead to a greater than necessary expenditure to the health care system. Hypothetical cost analysis of trigger finger injections have demonstrated the cost efficiency in performing at least 2 injections, but many surgeons continue to offer only a single injection.8
The purpose of this study was to use prospective data collected from patients presenting with STS to determine the efficacy of corticosteroid injections and to determine the approach with the most health care dollars saved when treating this condition. We hypothesized that offering up to 3 corticosteroid injections, rather than offering surgery after a failure of a first or a second injection, would be less expensive to payers for our cohort. The design of this study does not attempt to include all elements of cost, and does not take into account patient satisfaction, time out of work or travel, and inconvenience to patients. Its focus is on the money saved to the payer. This study was pragmatic in design, in that prospective data were collected without an attempt to randomize or encourage patients to pursue either injections or surgical management. Patients with recurrent symptoms were given the option for injection versus surgery. This approach was chosen because patients presenting with STS typically make the decision for surgery or injection based on a variety of factors including severity of symptoms, history of responsiveness to injection, occupation, or aversion to injections. We hypothesized that up to 3 injections would offer considerable health care dollar savings for the treatment of STS.
Materials and Methods
The institutional review board approval was obtained before enrollment of our patients in this prospective cohort study. Between March 2014 and September 2014, all patients presenting to a single surgeon at our institution with a diagnosis of STS were prospectively enrolled, and each affected digit was followed as a distinct event. Patients were excluded from this study if they refused injection, or if they had received any treatment in the digit to be studied for STS or another condition involving surgery in close proximity to the affected A1 pulley, such as Dupuytren’s contracture, carpal tunnel syndrome, De Quervain’s tenosynovitis, or a fracture.
All patients diagnosed underwent an initial corticosteroid injection by the senior author. Injections were performed with 1 mL (3 mg) of betamethasone and 1 mL of 1% lidocaine, infiltrated into the A1 pulley sheath. After injections, patients were counseled to return to the office on an as-needed basis with recurrent symptoms. No immobilization or therapy was prescribed. Patients were followed prospectively. On any return visit to the office, they were offered the option of either a repeat injection using the same technique, or surgical release. Enrollment in the study did not affect patient counseling or treatment options. Patients who did not return for subsequent visits were followed by telephone calls every 6 months, to ascertain if they had sought treatment elsewhere, or if they had any return of symptoms.
Injections were noted to be successful if the patients did not require any further treatment within the study period ending April 2016, and were considered failures if the patients required a repeat injection or surgery. Individual digits were injected a maximum of 3 times before being elected for surgery. Injection success rates were calculated for a first, second, or third treatment. The primary end point for follow-up was surgical release of the affected digit. If the patient had multiple fingers meeting enrollment criteria that were injected, these digits were followed individually, as other studies have described.
9
Calculations of cost savings to the payers were completed using the dollar amount reimbursed by insurers for initial office visits, return office visits, physician fee for injection, surgical facility fees, and surgeon fees. The dollar amount reimbursed, rather than the charge submitted to insurance companies, was used to best approximate the actual dollar amount spent to treat the patients. Reimbursements were calculated using actual fee schedules from the 6 insurance carriers serving patients in our cohort (Table 1). These included Medicare, Blue Cross/Blue Shield, United Healthcare, Tufts, Aetna, and Worker’s Compensation. The total reimbursed charges of treating our cohort were calculated. These costs were calculated additionally at each of 3 phases of care (Fig. 1). The first phase was an initial presentation for STS. The second phase of care was defined as those patients presenting with recurrent symptoms after a first injection. At the third phase of care, patients presented with persistent or recurrent symptoms after a second injection. The costs at each phase of care were then compared with the hypothetical cost of offering only surgery at each point. The average dollar amount saved per digit was calculated, as well as the percentage of total cost saved for the phase of care.
Table 1Reimbursement by Insurer
| Medicare ($) | Blue Cross/Blue Shield ($) | United Healthcare ($) | Worker’s Compensation ($) | Athena ($) | Tufts ($) | |
|---|---|---|---|---|---|---|
| Initial office visit (CPT 99203) | 112.85 | 118.62 | 148.66 | 141.99 | 120.12 | 145.37 |
| Subsequent office visit (CPT 99213) | 75.40 | 80.06 | 99.41 | 95.39 | 80.22 | 98.52 |
| Injection fee (CPT 20550) | 55.35 | 101.01 | 168.51 | 183 | 117.63 | 122.99 |
| Surgical facility fee | 327.83 | 820.99 | 697.59 | 813.76 | 608.32 | 733.31 |
| Surgical physician fee (CPT 26055) | 591.74 | 1,490 | 1,671.46 | 1,821.84 | 1,132.28 | 1,355 |
Figure 1Flowchart of our cohort’s treatment.
Statistical analysis involved the use of basic descriptive statistical formulas for both categorical and continuous data. In addition, a 1-way sensitivity analysis was used to determine the efficacy rate at which offering a third injection would not be cost effective (ie, the “inflection point”). In this analysis, the efficacy of the third injection was varied, whereas all other variables, including efficacies of first and second injections, cost of injections, and visits were held constant.
Results
A total of 82 patients were included in this study. Eighty-eight digits were injected and independently followed from enrollment to April 2016. The average follow-up for all digits was 21.9 months (range, 18.7–22.7 mo). No patients were lost to follow-up, and all patients treated successfully with injections were contacted every 6 months. None of these patients sought or received care elsewhere. Demographics, including diabetic status, are summarized in Table 2. Patients presented at a median of 2 months (range, 1 wk to 6 y) after the onset of symptoms. The average length of time between first and second injections was 140 days (72–243 d) and between the second and third injections was 161 days (31–426 d).
Table 2Demographics by Digit
| Gender | |
| Male | 29 digits (33%) |
| Female | 59 digits (67%) |
| Average age (range) | 62.4 y (28–93 y) |
| Diabetic | 21 digits (23.8%) |
| History of STS in other digit (%) | 26 digits (29.5%) |
Of the 88 digits initially injected, 40 did not require any further intervention throughout the follow-up period (45.4%). This represented the overall efficacy for the initial injection. Of the 48 digits requiring further intervention, 15 went on to surgery and 33 elected for another injection to the affected digit. Of the 33 patients who were injected a second time, 11 were successfully treated (33.3%), 15 went on to surgery, and 7 were injected a third time. The third injection was successful in 2 patients (28.6%), whereas surgery was indicated for the remaining 5 who did not improve to an acceptable level. Of the 128 injections administered, 53 injections resulted in the patient requiring no further treatment, representing an average individual injection efficacy of 41.4%, and a cumulative efficacy of 60.2% when offering up to 3 injections. Of the 5 patients enrolled with multiple trigger digits, 4 patients had 2 affected digits, and 1 patient had 3. For 3 of the 5 patients, only 1 digit required surgery during the 2-year follow-up, and for each of the other 2 patients, neither digit required surgery but only 1 digit required multiple injections. In the enrolled patient cohort, 77 follow-up office visits were billed, and 35 digits underwent surgical release. There were no complications associated with injection administration. Given that there were no complications in this series, we did not include costs of complications in our analysis.
The total reimbursement for treating our cohort was $96,358. This dollar amount includes visits, injections, and surgery. The hypothetical dollar amount that would have been spent to treat all patients with surgery at the first presentation was $169,089. Therefore, offering the first injection to all patients saved payers $72,731 ($826 per digit), or 43% of the total cost. Offering surgery as the only option to the 33 patients who failed a first injection would have cost the payers $70,231 including reimbursement for office visits and surgery. By offering a second injection, we calculated a cost savings for payers of $15,956 ($483 per digit) or 22.7% of the cost for the second phase of care. Offering a third injection saved $1,986 ($283 per digit) or 14.5% of the cost for the third phase of care. In calculating the total reimbursement savings to the payer, the total reimbursement of treatment for each digit was considered, including the reimbursement for each phase of care. Thus, patients who underwent 2 or 3 injections and went on to surgery cost more to their payer than patients who underwent a single injection before surgery. These data are summarized in Table 3.
Table 3Reimbursement Savings to Payer of Treatment With Corticosteroid Injection
| Phase of Care | Cost to Payer if Offering Injection ($) | Cost to Payer if Not Offering Injection ($) | Potential Total Savings ($) | Percentage of Potential Savings Per Phase of Care | Savings Per Patient ($) |
|---|---|---|---|---|---|
| Initial presentation (88 patients) | 96,358 | 169,088 | 72,731 | 43% | 826 |
| After failure of 1 injection (33 patients) | 54,274 | 70,231 | 15,956 | 22.7% | 484 |
| After failure of 2 injections (7 patients) | 11,719 | 13,706 | 1,986 | 14.5% | 284 |
∗ This cost accounts for all remaining patients undergoing surgery.
Given that the true efficacy of a third injection is unknown, and may vary depending on protocols or patient population, we conducted a sensitivity analysis to determine the minimum efficacy needed to maintain cost savings. The 1-way sensitivity analysis showed that, by fixing all other variables including first and second injection efficacy rates, complication rates, and the cost of all procedures/visits based on the proportion of insurances represented in our patients undergoing 3 injections, and varying the efficacy rate of the third injection, the “inflection point” of injection efficacy at which offering the third injection becomes cost-effective compared with offering all returning patients surgery is 10.28% (Fig. 2). Our efficacy rate for a third injection was higher, at 28.6%.
Figure 2Sensitivity analysis demonstrating the “inflexion point” (star) of third injection efficacy at which the third injection becomes cost-beneficial. The shaded area represents the region of potential cost savings based on the efficacy of the third injection. The cost of not offering a third injection was a constant value as all patients would have been offered surgery.
Discussion
The use of corticosteroid injections is a mainstay of the treatment of STS as the efficacy of a single injection at preventing the need for further treatment has recently been reported between 44% and 66%.
10
, 11
, 12
, 13
Although a recent investigation showed that a majority of providers only offer 1 injection,4
our study hypothesized and then demonstrated that offering 1, 2, and 3 injections instead of only surgery saved approximately 43%, 22.7%, and 14.5% of the cost at each treatment phase, respectively. Before this study, limited information existed to guide the most cost-effective algorithm for managing STS, and particularly the optimal number of injections to be given before offering surgery.In their study, Kerrigan and Stanwix’s
8
model used existing injection efficacy data in the literature to create a cost model for trigger injections. The data used for this study varied considerably from the findings within our cohort. The authors determined that the most cost-effective treatment was to attempt 2 injections before surgical release. However, their analysis was based on 3 studies that they stated have efficacy rates for a third injection of 0% to 5%.14
, 15
, 16
Therefore, the authors used a median 0% success rate in their calculations and concluded that a third injection has no benefit given its cost. However, the cited study by Benson et al14
showed an efficacy rate of 33% for patients undergoing a third injection, not 0%.15
In addition, the cited study by Newport et al16
concludes that 5% of patients required 3 or more injections to either achieve symptom remission or decide to pursue surgery, not that the efficacy of the third injection was 5%. Finally, more recent literature not used in the analysis of Kerrigan and Stanwix finds that third injections may be effective in up to 75% of cases.6
, 17
In our cohort, although the number of patients undergoing a third injection was small, we did not find a precipitous drop in efficacy for a third injection. The modest success rate of a third injection in our cohort (28.6%) was well above the “inflection point” determined in our sensitivity analysis (10.28%), thus resulting in a substantial cost-effectiveness (14.5% savings) when offered instead of surgery. Our sensitivity analysis and inflection point is similar to that described by Kerrigan and Stanwix (9%).Our patient sample was made up of 24% of diabetics. Within this subgroup, our data demonstrated a lower rate of surgery at 19% compared with nondiabetics at 54%, when treated with our protocol of offering up to 3 injections. Average treatment cost to payers for diabetic patients was $432, whereas it was $1167 for nondiabetic patients. Although this is counter to previous studies that suggest that diabetics respond more poorly to injections,
18
we suggest that a cost-effective approach for the payers would be to offer up to 3 injections irrespective of diabetic status.Our data are consistent with published data demonstrating limited morbidity of corticosteroid injections
19
and surgery.20
No complications from injections were noted in our cohort. Similarly, for the patients who underwent surgery, there were no revision A1 pulley releases, infections, digital nerve injuries, or other complications. If present, these surgical complications could have markedly increased the cost of an operative group of patients and enhanced the cost savings of offering an injection instead of surgery.This study has several limitations. The decision to repeat corticosteroid injections was subjective, and based on patient preference. The use of second and third injections was not based on a treatment algorithm as patients were allowed to choose to undergo repeat injections or to undergo surgery. This may potentially represent a bias in our data, given that patients with more severe symptoms may have been more likely to be resistant to injection and/or more likely to request surgery. Similarly, physician bias may have had an effect on treatment decisions, especially because of physician expectation of success or failure of an injection. In addition, all patients presenting with STS underwent at least 1 injection, and any patients who refused injections were excluded. This likely means that patients with a history of failed injections for other trigger digits, or those who had extremely severe symptoms, were not represented in this group. However, it is our practice to routinely offer injections as a first-line treatment even in patients with severe symptoms or who have had an A1 pulley release for a different digit in the past. Only 8 patients underwent a third injection, limiting the accuracy of our analysis of a third injection. However, our data regarding efficacy rates for a third injection are similar to recently published data that have a larger cohort receiving third injections.
17
Our study is also limited in that a single surgeon, using 1 protocol, performed all injections. A variety of injection techniques and protocols exist, and may vary in efficacy as well as cost.21
, 22
Furthermore, our cost estimates take into account the reimbursed dollar amounts that are specific to our institution, and these may vary by region, payer mix, and facility fees. Even so, our data are likely to underestimate the cost savings of treatment with injections instead of surgery as a result of our relatively low facility fees from the performance of procedures in an ambulatory surgery center rather than a full operating room. Surgery performed in an office-based setting or even percutaneous releases may represent an even lower cost that is represented in our data.23
Finally, although we believe that our follow-up period of 2 years is reasonable and represents the likely timeframe for recurrence, it may not be long enough to capture all instances of trigger finger recurrence after apparently successful injection, as up to16.4% of patients may require surgery or an additional injection greater than 2 years after their initial injection.13
In this study, no patients were lost to follow-up after undergoing injections, and none of these patients sought further care from a provider other than the senior author. It is possible that some patients who did not return after a trigger finger injection had persistent symptoms. From a cost perspective, none of these patients required any further billable medical treatment and therefore residual symptoms did not have an effect on our analysis, which solely focused on health care dollars spent by the payer.
On the basis of the possible cost savings to payers, this study represents a pragmatic cost analysis of performing multiple trigger finger injections before offering surgery for STS. We recommend that hand surgeons consider offering up to 3 corticosteroid injections before scheduling patients for surgical release. Although not every patient will be amenable to 3 injections, and the severity of symptoms or patient preference may make a second or third injection undesirable, the option of a second or third injection does offer a significant cost-savings benefit to the payer. Future investigation into the cost of care for STS should include a comprehensive analysis using postoperative disability time, lost earning potential, and patient outcomes, and could offer insight into overall value of injections versus surgery.
Supplementary Data
- Data Profile
References
- Cost analysis and related factors in patients with traumatic hand injury.J Hand Surg Eur Vol. 2013; 38: 673-679
- Treatment of ulnar neuropathy at the elbow: cost-utility analysis.J Hand Surg Am. 2012; 37: 1617-1629.e1613
- Utility and cost analysis of radiographs taken 2 weeks following plate fixation of distal radius fractures.J Hand Surg Am. 2015; 40: 1106-1109
- National utilization patterns of steroid injection and operative intervention for treatment of common hand conditions.J Hand Surg Am. 2016; 41: 367-373.e362
- A prospective randomized trial comparing the effectiveness of one versus two (staged) corticosteroid injections for the treatment of stenosing tenosynovitis.Hand (NY). 2014; 9: 340-345
- Management of stenosing flexor tenosynovitis: maximizing nonoperative success without increasing morbidity.Plast Reconstr Surg. 2016; 137: 557-562
- Treatment of flexor tenosynovitis of the hand ('trigger finger') with corticosteroids. A prospective study of the response to local injection.Arch Intern Med. 1991; 151: 153-156
- Using evidence to minimize the cost of trigger finger care.J Hand Surg Am. 2009; 34: 997-1005
- Long-term effectiveness of corticosteroid injections for trigger finger and thumb.J Hand Surg Am. 2015; 40: 121-126
- The efficacy of steroid injection in the treatment of trigger finger.Clin Orthop Surg. 2012; 4: 263-268
- Corticosteroid injection therapy for trigger finger or thumb: a retrospective review of 577 digits.Hand (NY). 2013; 8: 439-444
- Trigger finger: prognostic indicators of recurrence following corticosteroid injection.J Bone Joint Surg Am. 2008; 90: 1665-1672
- Long-term outcomes following a single corticosteroid injection for trigger finger.J Bone Joint Surg Am. 2014; 96: 1849-1854
- Injection versus surgery in the treatment of trigger finger.J Hand Surg Am. 1997; 22: 138-144
- Nonoperative treatment of trigger fingers and thumbs.J Hand Surg Am. 1989; 14: 553-558
- Treatment of trigger finger by steroid injection.J Hand Surg Am. 1990; 15: 748-750
- Long-term effectiveness of repeat corticosteroid injections for trigger finger.J Hand Surg Am. 2017; 42: 227-235
- Treatment of trigger finger in patients with diabetes mellitus.J Hand Surg Am. 1995; 20: 787-789
- Controlled study of the use of local steroid injection in the treatment of trigger finger and thumb.J Hand Surg Br. 1992; 17: 69-70
- Risk factors for complications of open trigger finger release.Hand (NY). 2015; 10: 297-300
- Ultrasound-guided first annular pulley injection for trigger finger.J Ultrasound Med. 2009; 28: 737-743
- Corticosteroid injections for trigger digits: is intrasheath injection necessary?.J Hand Surg Am. 1998; 23: 717-722
- Percutaneous trigger finger release: a cost-effectiveness analysis.J Am Acad Orthop Surg. 2016; 24: 475-482
Article Info
Publication History
Published online: June 08, 2018
Accepted:
April 9,
2018
Received:
May 29,
2017
Footnotes
No benefits in any form have been received or will be received related directly or indirectly to the subject of this article.
Identification
Copyright
© 2018 by the American Society for Surgery of the Hand. All rights reserved.
30874-2&id=gr1.jpg){kind=link}
30874-2&id=gr2.jpg){kind=link}