Trigger digit release: rates of surgery and complications as indicated by a United States Medicare database

Introduction

Surgical release of the A1 pulley is generally indicated in cases that have failed conservative measures or in instances where surgical management is the preferred treatment modality of the patient and/or surgeon (Lange-Riess et al., 2009; Lim et al., 2007; Patel and Bassini, 1992; Ryzewicz and Wolf, 2006; Thorpe, 1988; Turowski et al., 1997; Vaes et al., 1998). The literature has indicated that this procedure, whether performed open or percutaneously, is relatively simple and benign, with a low risk of complication or recurrence. Within existing literature, inconsistent definitions of complications and treatment failure, as well as significant variability in sample size and duration of post-operative follow-up, has led to substantially differing reported complication rates after trigger digit release (Gulabi et al., 2014; Huang et al., 2015; Lange-Riess et al., 2009; Lim et al., 2007; Thorpe, 1988).

The aim of this study was to use a national database to evaluate modern trends in surgical release of the A1 pulley for trigger digit, complication and revision rates and associated risk factors for complications and need for revision surgery.

Methods

The PearlDiver Database (www.pearldiverinc.com, Fort Wayne, IN), a fee-based publicly available insurance database, was utilized for data analysis. It contains records of patients with International Classification of Diseases, 9th Revision (ICD-9) diagnoses and procedures or Current Procedural Terminology (CPT) codes. All data were completely de-identified and anonymous and so was deemed exempt from Institutional Review Board approval. The data for this study were derived from a Medicare-based database within PearlDiver, which has over 100 million individual patient records from 2005–2012. The Medicare database was chosen as it includes CPT code modifiers, including those for laterality and affected digits. The patients in this database are tracked across all episodes of care in which their Medicare insurance is utilized. This includes primary episodes of care where the surgical procedure occurred, such as in the hospital, an ambulatory surgical centre or a private outpatient surgery centre, and any secondary episodes of health care, including encounters with their primary surgeon, another surgeon within the United States, a therapist that bills Medicare insurance or any practitioner in the primary care setting.

The database was queried for trigger digit release using CPT code 26055 (tendon sheath incision, e.g. for trigger digit). To assure that all included patients had a tendon sheath incision for a trigger digit and not for some other diagnosis; only those patients with CPT code 26055 associated with a diagnosis of trigger digit (ICD-9 727.03, trigger digit) were included. Trigger digit procedures were then sub-divided into the digit on which the operation was performed using CPT modifier codes as follows: left thumb (FA), left index finger (F1), left middle finger (F2), left ring finger (F3), left little finger (F4), right thumb (F5), right index finger (F6), right middle finger (F7), right ring finger (F8) and right little finger (F9).

Complications after trigger digit release, including finger stiffness and infection, were assessed for a 90-day postoperative period. A 90-day postoperative time window was chosen to reduce the risk that generic codes for postoperative complications were related to a subsequent surgical procedure. Complication codes could be added to a patients’ record by any practitioner that diagnosed the complication in either an outpatient or an inpatient setting. Table 1 describes the codes utilized for both complications. Complex regional pain syndrome and digital nerve injury were also assessed using ICD-9 codes, but insufficient patients with these diagnoses prevented these from inclusion in the final analysis. Revision trigger digit release was queried by searching for patients with a subsequent CPT code for trigger digit release (26055) on the same hand and digit as a previous trigger digit release. Revision trigger digit release within 1 year and 3 years of the index procedure was queried. Data for longer time periods were too small to be considered worth assessing.

OPEN IN VIEWER

Table 1.

ICD-9 and CPT codes used for postoperative complications.

Complication	ICD-9/CPT code	Description
Digit stiffness	ICD-9 718.44	Contracture of hand joint
	ICD-9 718.54	Ankylosis of hand joint
	ICD-9 719.54	Joint stiffness of hand, not classified elsewhere
Postoperative infection	CPT 20005	Incision and drainage of soft tissue abscess
	ICD-9 681.00	Unspecified cellulitis and abscess of finger
	ICD-9 681.9	Cellulitis and abscess of unspecified digit
	ICD-9 711.04	Pyogenic arthritis involving the hand
	ICD-9 711.94	Unspecified infective arthritis involving hand
	ICD-9 998.51	Infected postoperative seroma
	ICD-9 998.59	Other postoperative infection

We also tried to assess associations between various patient demographics, medical comorbidities, postoperative complications and revision surgery. Patient age groups (<65, 65–69, 70–74, 75–79, 80–84, 85+ years) and gender were recorded. Diabetes mellitus, peripheral vascular disease, liver disease, chronic kidney disease, rheumatoid arthritis, Dupuytren’s disease, obesity, morbid obesity and tobacco use were noted.

Statistical comparisons of cohort demographics and risk factors for complications were completed with Pearson χ² analysis. Odds ratios were calculated with respective 95% confidence intervals. For all statistical comparisons, p < 0.01 was considered significant.

Results

Trigger digit release

A total of 209,634 patients who underwent trigger digit release were included in the study. Overall, the number of trigger digit releases increased 16% when averaging the rate from 2005–2007 and comparing with 2010–2012 (P < 0.0001) (Table 2). The epidemiology and comorbidities of patients undergoing trigger digit release are provided in Table 2. There were increases in releases performed on each digit; however, the most substantial increases were noted in the middle and ring fingers (Figure 1). The middle finger was the most common digit to undergo trigger digit release, followed by the ring finger (P < 0.0001) (Figure 2). The little finger was least frequently released. Right hand trigger releases were more common than left hand (P < 0.0001).

OPEN IN VIEWER

Table 2.

Summary of epidemiologic data for trigger digit release patients, 2005–2012.

	2005	2006	2007	2008	2009	2010	2011	2012
Summary
Total number of points in database	24,971,819	25,655,246	25,749,249	25,615,694	25,375,062	25,626,128	26,101,098	26,505,506
Number of trigger releases	24,709	25,090	23,478	23,781	25,206	27,127	28,850	31,393
Percentage of trigger releases/year (%)	0.099	0.098	0.091	0.093	0.099	0.106	0.111	0.118
Demographics
Female (%)	71	69	67	66	65	64	64	64
Male (%)	29	31	33	34	35	36	35	35
Age <65 (%)	16	16	16	17	18	18	18	18
Age 65–69 (%)	27	26	25	26	26	27	28	28
Age 70–74 (%)	22	21	21	21	22	21	22	22
Age 75–79 (%)	18	18	18	17	16	16	16	16
Age 80–84 (%)	11	12	12	12	11	11	11	10
Age 85+ (%)	5.7	6.2	6.7	6.9	6.8	6.8	6.4	6.7
Tobacco use (%)	10	11	11	12	12	12	11	11
Obesity (%)	9.2	10	10	11	11	11	11	11
Morbid obesity (%)	5.7	5.3	5.9	6.3	6.8	6.9	6.7	6.4
Comorbidities
DM (%)	44	45	45	45	45	44	43	43
PVD (%)	16	17	16	16	15	15	14	14
CKD (%)	19	20	19	19	17	16	17	16
Liver (%)	6.7	6.3	6.6	7.1	6.9	6.8	6.7	6.5
RA (%)	6.4	6.4	7.0	6.8	6.7	6.5	6.1	5.8
Gout (%)	8.3	8.8	8.9	9.0	8.7	8.6	8.1	7.7
Dupuytren’s (%)	3.8	4.3	4.7	4.6	4.9	4.9	4.7	4.1

CKD: chronic kidney disease; DM: diabetes mellitus; Liver: chronic liver disease; PVD: peripheral vascular disease; RA: rheumatoid arthritis.

OPEN IN VIEWER

Figure 1.

Trends in trigger finger release in Medicare patients from 2005 to 2012. There was an increase in the incidence of release in all digits, but most pronounced in the middle and ring finger. The middle finger was the most frequent digit to undergo trigger finger release.

OPEN IN VIEWER

Figure 2.

Total number and laterality of trigger finger releases in Medicare patients from 2005 to 2012.

Risk factors for complications and revision

Dupuytren’s disease, rheumatoid arthritis, liver disease, obesity, tobacco use, peripheral vascular disease and diabetes mellitus were all associated with an increased rate of revision trigger digit release (Table 3). Of all demographics assessed, only <65 years was significantly associated with increased revision rates (Table 3). Gender and other age categories had no significant association with subsequent revision surgery.

OPEN IN VIEWER

Table 3.

Summary of Significant Risk Factors for Complications after Trigger Digit Release.

Revision surgery	Revised patients (%)	Non-revised patients (%)	OR (95% CI)	P
Diabetes	49	44	1.2 (1.1–1.4)	0.001
PVD	18	14	1.3 (1.1–1.5)	0.001
Liver disease	8.8%	6.4%	1.4 (1.2–1.7)	0.001
RA	8.4%	6.4%	1.4 (1.1–1.7)	0.004
Dupuytren’s	7.8%	4.3%	1.9 (1.5–2.3)	<0.0001
Tobacco	19	15	1.3 (1.1–1.5)	<0.0001
Obesity	17	14	1.3 (1.1–1.5)	0.004
Age <65	24	17	1.5 (1.3–1.7)	<0.0001
Stiffness	Stiff patients (%)	Non-stiff patients (%)	OR (95% CI )	P
Diabetes	47	44	1.2 (1.1–1.3)	0.003
PVD	17	14	1.2 (1.1–1.4)	0.001
RA	10	6.3	1.7 (1.5–2.0)	<0.0001
Dupuytren’s	9.5	4.2	2.4 (2.0–2.8)	<0.0001
Age ⩾85	9.1	6.7	1.4 (1.2–1.6)	<0.0001
Infection	Infected patients (%)	Non-infected patients (%)	OR (95% CI)	P
Diabetes	56	44	1.6 (1.4–1.9)	<0.0001
PVD	23	14	1.8 (1.6–2.2)	<0.0001
CKD	23	17	1.5 (1.3–1.8)	<0.0001
Liver disease	10	6.4	1.6 (1.3–2.1)	<0.0001
Dupuytren’s	6.6	4.3	1.6 (1.2–2.1)	0.002
Tobacco	27	15	2.0 (1.7–2.3)	<0.0001
Obesity	37	17	1.5 (1.2–1.8)	<0.0001
Morbid obesity	17	8.2	2.4 (2.0–2.8)	<0.0001
Age <65	37	17	2.9 (2.5–3.3)	<0.0001
Male gender	39	33	1.3 (1.1–1.5)	0.001

CI: confidence interval; CKD: chronic kidney disease; OR: odds ratio; PVD: peripheral vascular disease; RA: rheumatoid arthritis.

Patients with postoperative stiffness after trigger digit release had higher rates of Dupuytren’s disease, rheumatoid arthritis, an age greater than 85 years, diabetes mellitus and peripheral vascular disease (Table 3).

An age less than 65 years, morbid obesity, tobacco use, peripheral vascular disease, liver disease, Dupuytren’s disease, diabetes mellitus, chronic kidney disease, obesity and male were all associated with infection after trigger digit release (Table 3).

Discussion

The present study offers an assessment of current trends in trigger digit release. We noted a trend toward more operative trigger digit releases between 2005 and 2012. This is an interesting finding, as prior studies have demonstrated success rates between 61% and 84% for one to three corticosteroid injections in patients with new onset triggering (Clark et al., 1973; Fauno et al., 1989; Newport et al., 1990; Patel and Bassini, 1992; Shinomiya et al., 2016; Wojahn et al., 2014). The reason for this increase may be multifactorial, as patients may have better access to healthcare or are less inclined to pursue non-operative management given the very high rate of success of trigger digit release or the number of patients seeking treatment for the condition may also have increased over the study period (Cakmak et al., 2012; Lange-Riess et al., 2009; Ryzewicz and Wolf, 2006). Additionally, this increase could also be due to improved or more accurate coding for this procedure over the time period.

The rate of infection and wound complications following open trigger digit release are not currently well characterized. Lim et al. reported a series of 483 operative trigger releases and noted no instances of infection (Lim et al., 2007). Turowski et al. also noted no infections after trigger release in a series of 75 patients (Turowski et al., 1997). The results of these studies are misleading, as they are inadequately powered to report a true infection rate after trigger digit release. We found a 0.5%–0.6% rate of postoperative infection within 90 days after trigger release. While low, the rate of infection is not negligible. We also identified several important risk factors for postoperative infection, including diabetes mellitus, tobacco use, peripheral vascular disease, rheumatoid arthritis, Dupuytren’s disease, obesity and older age.

Loss of motion or persistent contracture following trigger digit release is another important complication that is also not well characterized, although reported rates are low or approaching 0% (Baumgarten, 2008; Bruijnzeel et al., 2012; Lange-Riess et al., 2009; Lim et al., 2007; Ryzewicz and Wolf, 2006; Thorpe, 1988; Turowski et al., 1997; Will and Lubahn, 2010). We found rates to be somewhat higher, ranging from 0.8% in the thumb to 1.6% in the index and middle fingers. Previously identified risk factors for postoperative stiffness or contracture include diabetes and male gender (Blyth and Ross, 1996; Bruijnzeel et al., 2012; Lange-Riess et al., 2009; Lim et al., 2007; Patel and Bassini, 1992; Strom, 1977; Turowski et al., 1997; Will and Lubahn, 2010). We found that diabetes mellitus, peripheral vascular disease, rheumatoid arthritis, Dupuytren’s disease, tobacco use, obesity and age less than 65 years were associated with postoperative stiffness.

Quantifying failure of trigger digit release and the rate of re-operation is challenging based on existing studies, as the definition of treatment success varies widely across studies (Lange-Riess et al., 2009; Lim et al., 2007; Thorpe, 1988). Bruijnzeel et al. reported 0.6% of patients had persistent triggering noted within 6 weeks of surgery in a study of 1707 separate trigger digit releases (Bruijnzeel et al., 2012). The authors correlated younger age and diabetes mellitus with recurrence and need for revision surgery. We found that the rate of revision varied based on the operative digit and that diabetes mellitus, Dupuytren’s disease, younger age, peripheral vascular disease, rheumatoid arthritis, tobacco use and obesity were risk factors for revision surgery.

This study has several limitations, many of which are common to studies that employ large insurance databases (Kandil et al., 2014; Werner et al., 2014; Yang et al., 2014; Yeranosian et al., 2013, 2014; Zhang et al., 2012, 2013, 2014). The power of the analysis is dependent on the accuracy and quality of the coded data, which includes proper entry of billing codes and non-coding by physicians all as potential sources of error. Errors in coding could lead to under- or over-reporting of complications in the present study. Furthermore, current coding does not allow differentiation of open trigger digit releases and percutaneous trigger digit releases that were coded as open releases. Additionally, the PearlDiver database currently only indexes an 8-year increment of data for the populations of interest. While the study period was large enough to find significant trends, some data may not be accurately captured in an 8-year time period. Additionally, the data reported includes only United States Medicare patients, limiting its generalizability to other healthcare systems. Despite our large patient numbers, we were unable to find adequate numbers of some complications, including complex regional pain syndrome or digital nerve injury to warrant reporting. Finally, the data are reported from the database in cohorts, which prevents multivariate or regression analysis, as the latter require line-by-line data for individual patients, not grouped (‘cohorts’) of data. The inability to perform multivariate analysis prevents evaluation of the independent effect of the evaluated variables on the risk of complications or revision following trigger digit release.

In conclusion, the incidence of trigger digit release in Medicare patients is increasing, with the middle digit the most frequently released. The rate of postoperative stiffness and infection are low. The need for revision within 3 years of initial trigger digit release ranges from 0.3% to 0.8%. Certain risk factors for complications exist, including diabetes, Dupuytren’s disease, smoking, rheumatoid arthritis, obesity and age.