Comparison of efficacy and safety of USG-guided versus blind pulley release for trigger finger: A systematic review and meta-analysis

Abstract

Background

To compare the efficacy and safety of the USG-guided vs blind pulley release (PR) for Trigger Finger by performing a meta-analysis of all relevant studies in the published literature.

Methods

A thorough and methodical search of the PubMed, Embase, Scopus, and Cochrane Library databases was carried out. Review Manager Software (RevMan) 5.4.1 was used to analyze the extracted data, and the results were displayed as forest plots with matching 95 % confidence intervals.

Results

The primary efficacy outcome i.e. residual triggering was significantly lower in USG-guided PR as compared to blind PR with a risk ratio of 0.16 (95 % CI 0.03–0.87), p = 0.03. The primary safety outcome i.e., percentage of complications was lower in the USG-guided procedure group, although the result was not statistically significant with a risk ratio of 0.25 (95 % CI 0.05–1.16), p = 0.08 with I² of 0 %. The operation time was longer in the USG-guided PR as compared to the blind procedure, although the difference was not statistically significant with a mean difference of 5.36 (95 % CI: −3.73, 14.46), p = 0.25. The postoperative visual analog scale (VAS) score at 4 weeks was significantly lower in USG-guided PR versus blind PR with a mean difference of −0.40 (95 % CI: −0.68, −0.33), p = 0.004.

Conclusion

When compared to blind PR, ultrasound-guided A1 PR for trigger finger was proven to be a safer and more economical method.

1. Introduction

"Trigger Finger (TF)," also referred to as "tendovaginitis" and "stenosing tenosynovitis," is brought on by the disproportionate sizing caused by either the thickened first annular pulley located at the head of the metacarpal region on the volar surface of the hand or a reaction in the synovial lining of the tendon sheath (TS) or the thickening of the tendon itself in the sheath in the form of a nodule, or a variable combination of these features, thereby hampering the smooth and unimpeded excursion of the flexor tendon.^{1, 2, 3, 4, 5, 6} 2.6 % of people in the general population have TF and it is more common in individuals with diabetes mellitus (DM), rheumatoid arthritis (RA), and middle-aged women. However, it can also affect children.^{7, 8, 9} There is a lot of confusion owing to the existence of parallel classification systems with overlapping grades in the literature, namely classic Quinnell grading, Modified Quinnell grading, and Green's classification of trigger fingers (Table 1).^{10, 11, 12} Patients with a TF typically experience pain, swelling, difficulty extending the finger, and a triggering sensation, all of which significantly impact daily routine activities.^13,14

Table 1.

Comparison of classification systems.

Grade Clinical findings (during flexion and extension)	The Quinnell grading of trigger finger10	Modified Quinnell grading of trigger finger11	Green classification of trigger finger12
0	Normal movement	–	–
I	Uneven movement	Normal movement, no pain	Palm pain and tenderness at A1 pulley, history of catching
II	Actively correctable	Normal movement, occasional pain	Demonstrable catching of digit, but can actively extend the digit
III	Passively correctable	Uneven movement involving crepitus or clicking without locking	Demonstrable locking of digit, but passively correctable
IV	Fixed deformity	Intermittent locking, actively correctable	Fixed, locked digit
V	–	Locking, only passively correctable.	–

In the literature, a myriad of treatment strategies for TF has been suggested including nonsteroidal anti-inflammatory drugs (NSAIDs) injection,¹⁵ steroid injection,^16,17 hyaluronic acid injection,¹⁸ physiotherapy,¹⁹ percutaneous release,²⁰ dry needling²¹ and open surgery.²² Recently, some randomized controlled trials (RCTs) examining the efficacy of TF release under ultrasonography (USG) guidance have been undertaken.^{23, 24, 25, 26, 27, 28} Two systematic reviews assessed the efficacy of various therapies for TF.^29,30 One of them compared the efficacy of steroid injection with other treatments and classified treatment plans into surgical and non-surgical.²⁹ Although other systematic reviews compared the effectiveness of various therapies, it was still unclear whether one was more beneficial than the other.³⁰ Another study sought to analyze the efficacy of various TF treatment strategies using pairwise meta-analysis,³¹ but it did not specifically mention the USG-guided technique as a sub-group.

No meta-analysis has been done so far comparing the efficacy and safety of the USG-guided vs blind pulley release (PR) for TFs from the published literature. Although existing literature is limited, the findings of the meta-analysis can help clinicians in selecting the optimal treatment approach for the management of TF requiring PR.

2. Materials and methods

The Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) guidelines were adhered to in the execution of the systematic review and meta-analysis.

The protocol was registered prospectively in PROSPERO (ID: CRD42023467900).

2.1. Search strategy

A thorough and methodical search of the literature was conducted in the aforementioned databases: PubMed, Embase, Scopus, and Cochrane Library. Using the thesaurus and free text forms for the essential concepts (TF, percutaneous pulley release, USG-guided, blind), two reviewers (RR, HK) separately carried out the search. To combine these concepts, we used the Boolean operators "OR" or "AND." The search approach was restricted to adult patients and human research, with English serving as the only language allowed. There was no constraint on the year or date.

Originally developed for the PubMed database, the search approach was subsequently modified for use in other databases (The detailed search strategy in the Supplementary file). The references of the included papers were checked to guarantee a thorough literature search, and citation monitoring of the included studies was carried out to find any more pertinent articles. After that, the studies were vetted by two separate reviewers (RR, HK), and any duplicates were eliminated.

Then, two reviewers (RR, HK) independently obtained and examined the entire text of the possibly eligible papers in accordance with the inclusion-exclusion criteria (Table 2). Whenever there was a disagreement, it was settled by consensus-building and discussion. If required, a third reviewer arbitrated.

Table 2.

PICOs framework of the studies in the Meta-analysis and the inclusion - exclusion criteria.

Key Concept	Inclusion Criteria	Exclusion Criteria
Population	Trigger Finger (TF) in adult patients	Children
Intervention	Ultrasonography (USG)-guided pulley release (PR)
Comparator	Blind PR as comparator or control group	•Open surgery •Drugs (like Local anesthetics, analgesics, corticosteroids, hyaluronic acid, or any other anti-inflammatory agent)
Outcomes	Primary Efficacy Outcome: •Residual Triggering (RT) Primary Safety Outcome: •Percentage of complications post-operatively Secondary Outcomes: •VAS Score at 4 weeks •Operation time
Type of Study	•Randomized controlled trials (RCTs) with control group •Non-randomized controlled studies (non-RCTs) •Retrospective analysis with control group	•In-vitro studies •Animal or Pre-clinical Studies •Cadaveric studies •Opinions and Editorials •Studies without a control group •Case reports •Case-control studies •Registries •Single arm studies with comparator arm as blind pulley release •Systematic reviews

2.2. Data extraction

Two reviewers (RR, HK) independently extracted the data from the included studies using a data extraction sheet in MS Excel: first author, year of publication, sample size, type of study, method of randomization, blinding and allocation concealment, details of intervention and control, primary and secondary outcomes, sample size of each arm, results of the study. The extracted data was checked for completeness and was analyzed using Review Manager Software (RevMan) 5.4.1.

2.3. Risk of bias assessment

Two impartial reviewers used the risk-of-bias assessment technique(ROB-2 tool) (https://www.riskofbias.info/welcome/rob-2-0-tool/current-version-of-rob-2) for randomized controlled studies to complete the assessment of potential bias. The instrument evaluated the risk of bias associated with each study in five areas: bias resulting from the randomization procedure, deviation from the planned interventions, incomplete outcome data, the outcome measurement, and selective result reporting.

Each individual studies was scored as "low risk of bias”, “some concerns” or “high risk of bias” across all domains.³² Observational studies (case-control or cohort study) were evaluated for quality using the Newcastle-Ottawa Scale (NOS), and any discordances were resolved with consensus or by involving a third reviewer.³³

2.4. Effect measures

The data analysis was done using RevMan 5.4.1 and results were presented in the form of forest plots with corresponding 95 % confidence intervals. Analysis was done considering the random effect model because of inherent heterogeneity in the type of studies contained in the meta-analysis. Heterogeneity testing was done and quantified using the I-square statistic. A statistically significant heterogeneity between the studies was defined as an I-square >25 %.

3. Results

The database search using the search strategy revealed 18 studies including 3 duplicates. Eight studies were found ineligible based on the title and abstract screening and the PRISMA flow graphic lists the exclusionary factors (Fig. 1). Full-text screening of the remaining seven studies was done using inclusion-exclusion criteria and four studies were further excluded from the analysis. Through the included papers' reference lists and citation screening, no new studies were found.

Fig. 1.

PRISMA flow diagram of the study screening.

3.1. Result synthesis

Overall, three studies, Lee et al.,²⁴ Muramatsu et al.,²⁵ and Pan et al.²⁶ were finalized in the meta-analysis and result synthesis evaluating a total of 94 patients undergoing USG-guided PR and 106 patients undergoing blind PR for the treatment of TF.

All three studies mentioned above were included in the primary outcome analysis using the random effect model.

3.2. Primary efficacy outcome

The primary efficacy outcome i.e., residual triggering (RT) was significantly lower in USG-guided PR as compared to blind PR with a risk ratio of 0.16 (95 % CI 0.03–0.87), p = 0.03 (Fig. 2). There was no heterogeneity or inconsistency in the studies included in terms of this outcome with an I² of 0 %.

Fig. 2.

Residual Triggering (RT) in USG-guided versus blind pulley release for Trigger Finger.

3.3. Primary safety outcome

The primary safety outcome i.e., percentage of complications was lower in USG-guided versus blind PR, although the result was not statistically significant with a risk ratio of 0.25 (95 % CI 0.05–1.16), p = 0.08 with I² of 0 % (Fig. 3). No nerve injury was reported in either of the two groups across included three studies.

Fig. 3.

Complications in USG-guided versus blind pulley release for Trigger Finger.

3.4. Secondary outcomes

The secondary outcomes that were analyzed were operation time and VAS Score at 4 weeks. The operation time was longer in the USG-guided PR versus blind PR, although the difference was not statistically significant with mean difference 5.36 (95 % CI: −3.73, 14.46), p = 0.25 with high heterogeneity; I² of 99 % (Fig. 4).

Fig. 4.

Operation time in USG-guided versus blind pulley release for Trigger Finger.

The postoperative visual analog scale (VAS) score at 4 weeks was significantly lower in USG-guided PR as compared to blind PR with mean difference −0.40 (95 % CI: −0.68, −0.33), p = 0.004 without any heterogeneity (I² of 0 %) in the studies included for analysis of this outcome (Fig. 5).

Fig. 5.

VAS Score in USG-guided versus blind pulley release for Trigger Finger.

3.5. Risk of bias assessment

An evaluation of the potential for bias in randomized controlled trials(RCTs) was conducted using the RoB2 assessment tool. Both RCTs, Lee et al. and Pan et al. were assessed for risk of bias for five domains as mentioned in the traffic light plot (Fig. 6). Lee et al. is a randomized study but unblinded and no allocation concealment was done in the study with overall assessment falling into the category of “some concerns” for bias (Fig. 7). Though Pan et al. mentioned the study as a randomized study but the method of randomization is not mentioned. In addition, allocation concealment and blinding were not done in this study with the overall assessment falling into the category of “high risk” of bias. Muramatsu et al. being a cohort study, the quality of study was assessed using the NOS scale and was found to be of moderate quality with NOS Score 6.

Fig. 6.

Traffic light plot for assessment of the risk of bias for the included RCTs using the RoB2 tool.

Fig. 7.

Risk of bias summary diagram of included randomized controlled trials Ts using the RoB2 tool.

3.6. Critical appraisal of individual studies

Muramatsu et al. is a methodologically different study in terms of being a retrospective cohort study with control cases (blind release) recruited from 2016 to 2018 and USG-guided procedure cases recruited from 2018 to 2020, whereas the other two studies Lee et al. and Pan et al. being RCTs. This itself could be a potential source of bias (study characteristics summarized in Table 3). Another difference is the inclusion of trigger thumb by Muramatsu et al. whereas the other two studies excluded trigger thumb cases.

Table 3.

Study characteristics summary of included studies.

Sr No	Author name	Type of trial	Methodology	Sample size (n)	Assessment of trigger score	Outcome	Results	Comments
1	Lee et al., 201824	RCT (2 arms: USG guided pulley release versus blinded pulley release)	Randomization was done using computer-generated sequence generation, Allocation concealment was not done, Blinding was not done	USG-guided: 23 Blind (control): 25	Quinnell Classification	•VAS score at 2 and 6 weeks •Residual triggering •Patient satisfaction	Significant improvement in VAS score, and post-operative contracture (p < 0.05) in USG-guided as compared to Blind trigger release. Residual triggering was 0 % in USG guided versus 3/25 in blind release. All patients who underwent USG-guided release were satisfied whereas in blind trigger release, three patients were not satisfied	Trigger thumb cases were excluded
2	Pan et al., 201926	RCT (2 arms: USG guided pulley release versus blinded pulley release)	Randomization, Allocation concealment was not done, Blinding was not done	USG-guided: 20 Blind (control): 21	Clinical semiquantitative evaluation criteria	•Clinical grade at day 7 •Residual triggering •Complications •Operation time	Significant improvement in clinical grade in USG-guided vs blind procedure at day 7 (p < 0.05) Triggering disappeared in all patients who underwent USG-guided pulley release whereas mild triggering continued in 15/21 patients and No Complications in USG-guided versus 2/21 complications in blind-trigger release. Mean Operation time was significantly longer in USG-guided (∼15.21 min) as compared to blind trigger release (∼5.23 min) (p < 0.05)	The method of randomization was not mentioned. Trigger thumb cases were excluded.
3	Muramatsu et al., 202125	Retrospective cohort study	A control group with limited blind trigger finger release was retrospectively included from 2016 to 2018. USG-guided Trigger finger release included from 2018 to 2020.	USG-guided: 51 Blind (control): 60	Green classification	•VAS score •DASH score •Residual triggering •Complications •Patient satisfaction (Patel and Moradia grading)	No difference between the two groups in terms of VAS score, operation time, and DASH score. No Residual triggering in USG-guided trigger finger release whereas 4/60 patients had residual triggering in blind trigger release No post-operative complications in USG-guided trigger finger release whereas 6/60 patients had post-operative complications.	Retrospective study with potential of bias. Trigger thumb cases were included.

Distinct tools have been employed in the participating studies, the Hanzhang needle knife (Beijing Huaxia Needle Knife Medical Equipment Factory) by Pan et al., the HAKI knife (BK Meditech Inc., Seoul, Korea) by Lee et al., and the Yasunaga knife (Medical U&A, Inc., Japan) by Muramatsu et al. The intra-synovial approach has been used by Muramatsu et al., while in the other two studies, it has been extra-synovial.

We feel a thorough understanding of the different classification systems of the TF is a prerequisite before embarking on a correct interpretation of observations and making a comparison (Table 1). A patient with uneven and slightly restricted finger movements where the classic triggering has not started as yet will be a grade I of classic Quinnell grading,¹⁰ grade III of modified Quinnell grading,^11, and grade II of Green's classification.¹² A fixed locked digit that cannot be extended even with passive manipulation does not find any place to be graded as per modified Quinnell grading, while it falls under grade IV of both classic Quinnell grading and Green's classification. Grade II of classic Quinnell grading and Green's classification is the same as grade IV of modified Quinnell grading. It won't be wrong to interpret that classic Quinnell grading is very similar to Green's classification with just an added grade 0 for the normal finger.

Pan et al. have mentioned that they have followed clinical semi-quantitative evaluation criteria based on Lapègue et al.³⁴ and Lee et al. But Pan et al. have actually used Lapègue et al., 2016 criteria; which states grade I as intermittent mild triggering (not continuous) whereas Lee et al. themselves followed classic Quinnell grading, 2018 which classifies uneven finger movements without proper triggering to be grade I. Muramatsu et al. on the other hand followed the Green classification for TF. These differences could be a potential reason for heterogeneity between the studies and should be kept in mind while interpreting the results.

Few researchers argue that it is the release of the proximal portion of the pulley that matters and that clinical outcomes may not always correlate with the distal part of the pulley. This may have an impact on the analysis or interpretation of results as no RT or discomfort may be seen despite an incomplete release.³⁵

3.7. Publication bias assessment

Detailed assessment was not made owing to only three studies, however the publication bias assessment has been presented using Funnel plot (Supplementary Fig. 1).

4. Discussion

In this meta-analysis, USG-guided PR showed significantly lower residual triggering rates (primary efficacy outcome) as compared to blind PR. The USG-guided PR group had a longer operation duration and a lower percentage of complications (primary safety outcome), but the disparateness was not statistically significant. No advantage of either technique could be observed in terms of improvement in VAS score.

While offering a complete, definite, and under-vision release of the A1 pulley, the conventional open surgical procedure is fraught with longer recovery periods, scarring and infection at the skin incision site especially in diabetics, and nerve injuries.^36,37 Though obviating the above-mentioned complications, the blind percutaneous procedure has its disadvantages in the form of incomplete release, residual triggering, injury to the digital nerve and vessel, and scoring of the underlying flexor tendon. USG-guided percutaneous release of the A1 flexor tendon pulley minimizes the disadvantages of the blind percutaneous procedure while simultaneously providing an early return to work and better cosmetic results²³ but with an added cost. Additionally, USG guidance allows for the identification and exclusion of secondary causes contributing to trigger fingers, thereby enhancing diagnostic accuracy.

The sonographic anatomy of the pulleys has been studied in the literature in quite detail^38,39 and the effectiveness of USG-guided percutaneous release of A1 pulley has also been demonstrated by independent research.^{40, 41, 42} and a recent systematic review.⁴³ Consequently, there has been a noticeable increase in the adoption of USG-guided release techniques over time.

Due to its oblique path across the MCP joint, the thumb's radial digital nerve is most vulnerable to injury in blind or non-image-guided releases. As they pass obliquely across the A1 pulley, the second most susceptible nerves are the ulnar digital nerve of the little finger and the radial digital nerve of the index finger. In contrast, the digital nerves in the middle and ring fingers travel lengthwise and away from the pulley, making the release of the A1 pulley in these fingers safe.⁴⁴ None of the three papers that make up this meta-analysis mentioned any cases of nerve damage.

Chopin C et al. evaluated USG-guided A1 pulley release and concluded that the length of duration of symptoms before the procedure could be an important covariate predicting the efficacy outcome, but the study lacked the control group.⁴² Other important covariates that can influence the outcome are the size of the pulley, sonographic characteristics, comorbidities like DM, and the number of injections of corticosteroids received before or during the release procedure.⁴⁵ Unfortunately, none of the papers included in our review evaluated these factors.

A meta-analysis by Zhao JG et al., 2014 was in line with our study and evaluated the overall success rate of percutaneous PR according to instrument type, cortisone use, and sonographic guidance. This study also highlighted that the patients who underwent the ultrasonography-guided release group had higher success rates ranging from 91 to 100 %. However, the point to note is that all four studies that evaluated sonographic guided PR incorporated in the meta-analysis by Zhao et al. were non-comparative, without any control group.⁴⁶ A recently published cohort study by Yavari et al., 2023 compared the percutaneous USG-guided release with classic open surgery, and the results are in similar line with faster recovery and less pain intensity in the USG-guided group as compared to the classic open-release group.²⁸

Recently, some concerns have been raised in the literature regarding new-onset post-opertaive Dupuytren Contracture after open surgical TF release.⁴⁷ This further emphasizes the importance of the USG-guided percutaneous release of the A1 pulley in the TF to minimize such complications.

4.1. Limitations

Certain limitations of this meta-analysis should be duly considered while interpreting the results. Firstly, there were only three studies relevant to our inclusion/exclusion criteria, with a low sample size. The point estimates I² or the heterogeneity statistic can be erroneous and biased in a small meta-analysis in which few studies have been included. It should be interpreted cautiously supplemented confidence intervals.⁴⁸ Secondly, lack of blinding and allocation concealment added to substantial possibility of bias in the included trials. One of the studies included was a retrospective cohort study, adding to the heterogeneity. Thirdly, there was a difference between the shape and size of the blade and the classification system used for grading trigger finger in all three studies, which could have easily influenced the outcomes in terms of completeness of PR and the final score. These variables should be considered while interpreting the results of the meta-analysis.

Adequately powered methodologically well-planned RCTs are required to provide conclusive evidence.

5. Conclusion

In conclusion, USG-guided A1 pulley release in TF was observed to be linked with a statistically significant lower incidence of residual triggering and a lower rate of complications as compared to blind pulley release. Additionally, patients undergoing USG-guided procedures experienced significantly greater improvements in VAS scores. However, the operation time for USG-guided procedures was longer as compared to blind percutaneous procedures. Overall, the USG-guided release of the A1 pulley in TF offers a safer, and cost-effective option, with better efficacy and fewer complications. Ultrasound guidance allows for real-time visualization and precise targeting of the A1 pulley and reduces the risk of inadvertent damage to the surrounding tissues such as nerves, tendons, and blood vessels.

Informed consent

Not applicable as the result was synthesized from already published studies.

Credit author statement

Rachna Rohilla: Methodology, Protocol writing, Literature search, Data extraction, Data analysis, Risk of bias assessment, Manuscript writing and editing.

Harmeet Kaur: Conceptualisation, Methodology, Protocol registration, Literature search, Data verification, Validation, Manuscript writing and Intellectual input.

Punit Tiwari: Data verification, Validation, Manuscript writing, and Intellectual input.

Presentation at a meeting

None.

Artificial intelligence

Artificial intelligence (AI) or AI-assisted technologies were not used in the preparation of the manuscript.

Ethical statement

The manuscript has been read and approved by all the authors, the requirements for authorship have been met, and each author believes that the manuscript represents honest work.

Funding

None.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following are the Supplementary data to this article:

figs1.