Injury, Int. J. Care Injured 45 (2014) 1889–1895
Contents lists available at ScienceDirect
Injury journal homepage: www.elsevier.com/locate/injury
Ulnar styloid process nonunion and outcome in patients with a distal radius fracture: A meta-analysis of comparative clinical trials M.M.E. Wijffels a,*, J. Keizer a, G.A. Buijze b, Y. Zenke c, P. Krijnen a, N.W.L. Schep d, I.B. Schipper a a
Department of Surgery-Traumatology, Leiden University Medical Center, Leiden, The Netherlands Department of Orthopedic Surgery, Academic Medical Center, Amsterdam, The Netherlands Department of Orthopaedic Surgery, University of Occupational and Environmental Health, Fukuoka, Japan d Department of Surgery, Academic Medical Center Amsterdam, Amsterdam, The Netherlands b c
A R T I C L E I N F O
A B S T R A C T
Article history: Accepted 5 August 2014
Purpose: There is no consensus on the relation between ulnar styloid process nonunion and outcome in patients with distal radius fractures. The aim of this study was to analyze whether patient-reported outcome is influenced by the nonunion of the accompanying ulnar styloid fracture in distal radius fracture patients. Methods: A meta-analysis of published studies comparing outcomes after distal radius fractures with a united versus a non-united ulnar styloid process was performed. In addition, if provided by the authors, the raw data of these studies were pooled and analysed as one study. The outcome measures of the analyses included patient-reported outcome, functional outcome, grip-strength, pain, and distal radioulnar joint (DRUJ) instability. Results: Data from six comparative studies were included, concerning 365 patients with a distal radius fracture. One hundred and thirty-five patients with an ulnar styloid union were compared with 230 patients with a nonunion of the ulnar styloid. No significant differences were found between groups regarding any outcome measure. Conclusion: Based on this meta-analysis, there is no relation between the nonunion of the ulnar styloid process and function in patients with a distal radius fracture. ß 2014 Elsevier Ltd. All rights reserved.
Keywords: Distal radius fracture Nonunion Ulnar styloid process Wrist fracture Distal radioulnar joint instability Outcome
Introduction One in six patients who present with a fracture at the emergency department has a distal radius fracture (DRF) [1–3]. Of those patients, 44–65% also have an ulnar styloid process (USP) fracture [4–7]. Displacement of DRFs may result in disruption of the distal radioulnar ligaments, either by direct ligamentous tear or by avulsion of the origin of the ligaments, that is, a fracture at the ulnar fovea or the ulnar styloid [5,6,8–17]. Ulnar styloid avulsion may result in distal radioulnar joint (DRUJ) instability due to triangular fibrocartilage complex (TFCC) disruption, the principal stabilizer of the DRUJ [9,18–22].
* Corresponding author at: Leiden University Medical Center, Department of Surgery-Traumatology, P.O. Box 9600, 2300 RC Leiden, The Netherlands. Tel.: +31 71 526 9111; fax: +31 71 526 6750. E-mail addresses: [email protected], [email protected] (M.M.E. Wijffels). http://dx.doi.org/10.1016/j.injury.2014.08.007 0020–1383/ß 2014 Elsevier Ltd. All rights reserved.
Nonunion of the USP fracture has been proposed as a potential source of ongoing ulnar-sided wrist pain and DRUJ instability [9,22,23]. Mentzel and colleagues reported a better residual function if the USP fracture unites [24]. Others found that nonunion of the ulnar styloid did not affect functional outcome or pain following volar plate fixation of a DRF [25–30]. To date, the relation between USP nonunion and outcome in patients with DRFs is unclear, due to the conflicting results and small sample sizes of the published studies. Therefore, a metaanalysis was performed to provide the best available evidence. The primary aim of this meta-analysis was to compare functional outcome between patients with and without the union of an ulnar styloid and a healed DRF. Materials and methods The meta-analysis was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) statement [31].
1890
M.M.E. Wijffels et al. / Injury, Int. J. Care Injured 45 (2014) 1889–1895
Types of studies and participants
Data collection
All studies comparing DRF patients, with and without union of the USP fracture, were assessed for inclusion in the present study.
The two surgical trainees (M.M.E.W. and J.K.) independently extracted data from all eligible studies. The corresponding authors of included papers were contacted by e-mail and requested to provide the raw data.
Inclusion and exclusion criteria Inclusion criteria were (1) English or German language, (2) reporting patient-reported outcome measure (PROM), pain, function, and/or DRUJ instability, (3) operatively or non-operatively treated patients, and (4) follow-up of 6 months or more. Exclusion criteria were (1) case series, (2) not available as full text article, (3) no reports on outcome data required for analyses, (4) no reports on 95% confidence intervals or standard deviations for outcome data, (5) the required data for this meta-analysis could not be derived from the published data, nor from the raw data, and (6) multi-trauma patients. Outcome measures The primary outcome of this meta-analysis was patientreported outcome based on questionnaires. Secondary outcome measures were range of motion, grip strength, pain, and clinical DRUJ instability. Furthermore, data extracted from the studies included patient characteristics, study design, sample size, type of radius fracture, length and completeness of follow-up, and type of treatment of the DRF. Search methods for identification of studies The systematic search was performed by a professional librarian and a member of our research team (M.M.E.W.) without any limit in publication date in PubMed, EMBASE, Web of Science, COCHRANE, CINAHL, Academic Search Premier, ScienceDirect, Springer, Wiley Blackwell, LWW, HighWire, and Informahealth. The search was conducted on October 22 2013. The search strategy combined various terms and synonyms for ulnar styloid fractures and DRFs. The complete search strategy is shown in the Appendix. Selection of studies After removing duplicates, two surgical trainees (M.M.E.W. and J.K.) screened the title and abstract of all studies identified in the initial search, using the inclusion and exclusion criteria. If no abstract was available, the full text was assessed for eligibility. Nonrelevant studies were excluded. Full text copies of the remaining studies were independently assessed according to the predefined inclusion and exclusion criteria. In addition, the reference lists of selected articles were screened for relevant studies not identified in the search, using the above criteria. Disagreements were resolved by means of discussion and arbitration by an experienced trauma surgeon (I.S.).
Analysis of raw data The available raw data were assembled in one database and analysed as if they were results from one study. Scores on patientreported outcome questionnaires, range of motion and grip strength (expressed as percentage of that of the uninjured wrist), and scores on a visual analogue scale for pain were compared between the union and nonunion groups using the Mann–Whitney test. Meta-analysis of the primary outcome In the selected studies, (a combination of) different questionnaires were used to measure patient-reported outcome. If different questionnaires were used within the same study, these data were pooled in the following way, in order to obtain one summary estimate for the PROMs per study. The results for the different questionnaires were adjusted so that they reflected a scale ranging from 0 (maximal disability) to 100 (no disability). Then, the standardised mean difference with 95% confidence interval within each study was calculated using the inverse variance method with fixed effects. These standardised mean differences from the individual studies were meta-analysed using the method of mean differences with random effects. Meta-analysis of the secondary outcomes Pronation, supination, flexion, and extension were analysed separately and reported as an absolute range of motion in degrees. Grip strength of the injured wrist was reported as a percentage of the uninjured wrist and could be pooled between studies without further adjustments. Other secondary outcome measures at follow-up included the percentage of patients with pain and the percentage of patients with DRUJ instability determined by clinical tests. For meta-analysis of the secondary outcomes, means with standard deviations or numbers (%) were extracted from the articles. The meta-analyses for continuous outcome were performed using the method of mean differences with random effects. The meta-analyses for dichotomous outcome were performed using the Mantel–Haenszel method with random effects. Differences in outcome between the union and nonunion groups with p < 0.05 were considered statistically significant. The Cochran’s chi-squared test (with significance defined as p < 0.1) and I2-index (with substantial heterogeneity defined as value >50%) were used to evaluate heterogeneity between studies. Results
Assessment of methodological quality Literature search The methodological quality of the selected studies was assessed using the quality assessment tool derived from the Grading of Recommendation Assessment, Development, and Evaluation (GRADE) Working Group [32]. The overall quality of evidence was based on study design, study quality, consistency, and directness. To evaluate the limitations in methodological quality, loss to follow-up and consistency of treatment strategy were scored. Disagreements were resolved by means of discussion, with arbitration by an independent trauma surgeon with a master in clinical epidemiology (N.S).
The search resulted in 488 potentially eligible studies, of which 17 full text articles were analysed for eligibility (Fig. 1). Eight studies met our inclusion criteria; however, two studies did not report sufficient data to be included in the meta-analysis leaving six studies available for meta-analysis [24–30]. These studies included a total of 135 patients in the union group and 230 in the nonunion group. Raw data were available for three out of the six studies [26,29,30]. Six patients (three in the union group and three in the
M.M.E. Wijffels et al. / Injury, Int. J. Care Injured 45 (2014) 1889–1895
1891
Idenficaon
References retrieved from data-bases PubMed, EMBASE, Web of Science, COCHRANE, CINAHL, Academic Search Premier, ScienceDirect, Springer, Wiley Blackwell,LWW, HighWire and Informahealth N=488 (aer removing duplicates)
References excluded Did not meet inclusion criteria. N=471 -19 language -41 case report -47 no abstract -16 chapters -3 comments Screening
-2 only abstract -2 animal studies -341 different subject
Eligibility
References Aer screening tles and abstracts N=17
Arcles excluded Did not meet inclusion criteria. N=9 -8 cohort studies
References
-1 no data on union-non union
Aer reading full-text arcle N=8 Arcles excluded No useful informaon N=2 -2 insufficient data
Included
References Aer reading full-text arcle N=6
Fig. 1. Study flow diagram of the systematic review.
nonunion group) had participated in both the studies of Buijze and Wijffels; their data were included only once in our meta-analysis. Description of included studies and outcome measure reporting The detailed description, reported outcomes, and methodological quality, based on GRADE guidelines, of the included studies in the meta-analysis are summarised in Tables 1 and 2. Analysis of raw data The combined raw data of the three studies [26,29,30] included the outcome data of 40 patients with union and 65 patients with
nonunion of the USP. No statistically significant differences between the combined union and nonunion groups were found for patientreported outcomes, range of motion, grip strength, and pain (Table 3). Meta-analysis: primary outcome Patient-reported outcome was measured in four studies [26,27,29,30] using (a combination of) different questionnaires: the Mayo wrist score [33] in three studies [26,27,30], the Gartland and Werley score [34] in two studies [26,30], and the Disabilities of the Arm, Shoulder and Hand (DASH) score [35] in three studies [26,27,29].
M.M.E. Wijffels et al. / Injury, Int. J. Care Injured 45 (2014) 1889–1895
1892 Table 1 Characteristics and results of the included studies. Study
n unions
n nonunions
Treatment of distal radius fracture
Mean age (range)
Analyzed outcome used in this meta-analysisb
Mean follow-up (range) (months)
Significant differences between groups
Buijze 2010a Kim 2011 Kraemer 2013
16 20 55
20 71 46
54 (25–75) 50 (18–88) 56 (18–81)
1,2,3,4,5,6,7 1,2,4,6,7 6,7
6 23 (12–48) 20 (6–49)
None None None
Mentzel 2001
6
29
59
6
12
Wijffels 2011a
16
18
Operative Operative Operative and conservative Operative and conservative Operative
56 (25–81)
2,3,4,5,6,7
31 (6–74)
22
46
Operative
64 (25–94)
1,4,5,6
14.9 (6–38)
Better range of motion in the union-group Better grip strength in the non-union group None
a
Zenke 2009 a b
Raw data available for analysis. (1) DASH, (2) Mayo Wrist Score, (3) Gartland and Werley score, (4) range of motion, (5) grip, (6) pain, (7) DRUJ instability.
Table 2 GRADE quality assessment of the included studies. Study a
Buijze 2010 Kim 2011 Kraemer 2013 Mentzel 2001 Wijffels 2011a Zenke 2009a a b c d e f g h i j k
Significant differences between groups
Study design
Inconsistencyb
Indirectnessc
Imprecisiond
Overall quality
None None None Better range of motion in the union group Better grip strength in the nonunion group None
Observational Observational Observational Observational Observational Observational
No No No No Seriouse No
No No Seriousj Seriousk Seriousi No
Seriousf No Seriousg Serioush Seriousf No
Low Low Very low Very low Very low Low
Raw data available for analysis. Inconsistency refers to unexplained heterogeneity of results. Indirectness refers to an indirect comparison of groups (e.g., loss to follow up and treatment strategy). Imprecision refers to studies with relatively few patients and few events and thus with wide confidence intervals around the estimate of the effect. No clear explanation for difference in grip strength. No DRUJ instability present. Only one questionnaire to include.
Contents lists available at ScienceDirect
Injury journal homepage: www.elsevier.com/locate/injury
Ulnar styloid process nonunion and outcome in patients with a distal radius fracture: A meta-analysis of comparative clinical trials M.M.E. Wijffels a,*, J. Keizer a, G.A. Buijze b, Y. Zenke c, P. Krijnen a, N.W.L. Schep d, I.B. Schipper a a
Department of Surgery-Traumatology, Leiden University Medical Center, Leiden, The Netherlands Department of Orthopedic Surgery, Academic Medical Center, Amsterdam, The Netherlands Department of Orthopaedic Surgery, University of Occupational and Environmental Health, Fukuoka, Japan d Department of Surgery, Academic Medical Center Amsterdam, Amsterdam, The Netherlands b c
A R T I C L E I N F O
A B S T R A C T
Article history: Accepted 5 August 2014
Purpose: There is no consensus on the relation between ulnar styloid process nonunion and outcome in patients with distal radius fractures. The aim of this study was to analyze whether patient-reported outcome is influenced by the nonunion of the accompanying ulnar styloid fracture in distal radius fracture patients. Methods: A meta-analysis of published studies comparing outcomes after distal radius fractures with a united versus a non-united ulnar styloid process was performed. In addition, if provided by the authors, the raw data of these studies were pooled and analysed as one study. The outcome measures of the analyses included patient-reported outcome, functional outcome, grip-strength, pain, and distal radioulnar joint (DRUJ) instability. Results: Data from six comparative studies were included, concerning 365 patients with a distal radius fracture. One hundred and thirty-five patients with an ulnar styloid union were compared with 230 patients with a nonunion of the ulnar styloid. No significant differences were found between groups regarding any outcome measure. Conclusion: Based on this meta-analysis, there is no relation between the nonunion of the ulnar styloid process and function in patients with a distal radius fracture. ß 2014 Elsevier Ltd. All rights reserved.
Keywords: Distal radius fracture Nonunion Ulnar styloid process Wrist fracture Distal radioulnar joint instability Outcome
Introduction One in six patients who present with a fracture at the emergency department has a distal radius fracture (DRF) [1–3]. Of those patients, 44–65% also have an ulnar styloid process (USP) fracture [4–7]. Displacement of DRFs may result in disruption of the distal radioulnar ligaments, either by direct ligamentous tear or by avulsion of the origin of the ligaments, that is, a fracture at the ulnar fovea or the ulnar styloid [5,6,8–17]. Ulnar styloid avulsion may result in distal radioulnar joint (DRUJ) instability due to triangular fibrocartilage complex (TFCC) disruption, the principal stabilizer of the DRUJ [9,18–22].
* Corresponding author at: Leiden University Medical Center, Department of Surgery-Traumatology, P.O. Box 9600, 2300 RC Leiden, The Netherlands. Tel.: +31 71 526 9111; fax: +31 71 526 6750. E-mail addresses: [email protected], [email protected] (M.M.E. Wijffels). http://dx.doi.org/10.1016/j.injury.2014.08.007 0020–1383/ß 2014 Elsevier Ltd. All rights reserved.
Nonunion of the USP fracture has been proposed as a potential source of ongoing ulnar-sided wrist pain and DRUJ instability [9,22,23]. Mentzel and colleagues reported a better residual function if the USP fracture unites [24]. Others found that nonunion of the ulnar styloid did not affect functional outcome or pain following volar plate fixation of a DRF [25–30]. To date, the relation between USP nonunion and outcome in patients with DRFs is unclear, due to the conflicting results and small sample sizes of the published studies. Therefore, a metaanalysis was performed to provide the best available evidence. The primary aim of this meta-analysis was to compare functional outcome between patients with and without the union of an ulnar styloid and a healed DRF. Materials and methods The meta-analysis was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) statement [31].
1890
M.M.E. Wijffels et al. / Injury, Int. J. Care Injured 45 (2014) 1889–1895
Types of studies and participants
Data collection
All studies comparing DRF patients, with and without union of the USP fracture, were assessed for inclusion in the present study.
The two surgical trainees (M.M.E.W. and J.K.) independently extracted data from all eligible studies. The corresponding authors of included papers were contacted by e-mail and requested to provide the raw data.
Inclusion and exclusion criteria Inclusion criteria were (1) English or German language, (2) reporting patient-reported outcome measure (PROM), pain, function, and/or DRUJ instability, (3) operatively or non-operatively treated patients, and (4) follow-up of 6 months or more. Exclusion criteria were (1) case series, (2) not available as full text article, (3) no reports on outcome data required for analyses, (4) no reports on 95% confidence intervals or standard deviations for outcome data, (5) the required data for this meta-analysis could not be derived from the published data, nor from the raw data, and (6) multi-trauma patients. Outcome measures The primary outcome of this meta-analysis was patientreported outcome based on questionnaires. Secondary outcome measures were range of motion, grip strength, pain, and clinical DRUJ instability. Furthermore, data extracted from the studies included patient characteristics, study design, sample size, type of radius fracture, length and completeness of follow-up, and type of treatment of the DRF. Search methods for identification of studies The systematic search was performed by a professional librarian and a member of our research team (M.M.E.W.) without any limit in publication date in PubMed, EMBASE, Web of Science, COCHRANE, CINAHL, Academic Search Premier, ScienceDirect, Springer, Wiley Blackwell, LWW, HighWire, and Informahealth. The search was conducted on October 22 2013. The search strategy combined various terms and synonyms for ulnar styloid fractures and DRFs. The complete search strategy is shown in the Appendix. Selection of studies After removing duplicates, two surgical trainees (M.M.E.W. and J.K.) screened the title and abstract of all studies identified in the initial search, using the inclusion and exclusion criteria. If no abstract was available, the full text was assessed for eligibility. Nonrelevant studies were excluded. Full text copies of the remaining studies were independently assessed according to the predefined inclusion and exclusion criteria. In addition, the reference lists of selected articles were screened for relevant studies not identified in the search, using the above criteria. Disagreements were resolved by means of discussion and arbitration by an experienced trauma surgeon (I.S.).
Analysis of raw data The available raw data were assembled in one database and analysed as if they were results from one study. Scores on patientreported outcome questionnaires, range of motion and grip strength (expressed as percentage of that of the uninjured wrist), and scores on a visual analogue scale for pain were compared between the union and nonunion groups using the Mann–Whitney test. Meta-analysis of the primary outcome In the selected studies, (a combination of) different questionnaires were used to measure patient-reported outcome. If different questionnaires were used within the same study, these data were pooled in the following way, in order to obtain one summary estimate for the PROMs per study. The results for the different questionnaires were adjusted so that they reflected a scale ranging from 0 (maximal disability) to 100 (no disability). Then, the standardised mean difference with 95% confidence interval within each study was calculated using the inverse variance method with fixed effects. These standardised mean differences from the individual studies were meta-analysed using the method of mean differences with random effects. Meta-analysis of the secondary outcomes Pronation, supination, flexion, and extension were analysed separately and reported as an absolute range of motion in degrees. Grip strength of the injured wrist was reported as a percentage of the uninjured wrist and could be pooled between studies without further adjustments. Other secondary outcome measures at follow-up included the percentage of patients with pain and the percentage of patients with DRUJ instability determined by clinical tests. For meta-analysis of the secondary outcomes, means with standard deviations or numbers (%) were extracted from the articles. The meta-analyses for continuous outcome were performed using the method of mean differences with random effects. The meta-analyses for dichotomous outcome were performed using the Mantel–Haenszel method with random effects. Differences in outcome between the union and nonunion groups with p < 0.05 were considered statistically significant. The Cochran’s chi-squared test (with significance defined as p < 0.1) and I2-index (with substantial heterogeneity defined as value >50%) were used to evaluate heterogeneity between studies. Results
Assessment of methodological quality Literature search The methodological quality of the selected studies was assessed using the quality assessment tool derived from the Grading of Recommendation Assessment, Development, and Evaluation (GRADE) Working Group [32]. The overall quality of evidence was based on study design, study quality, consistency, and directness. To evaluate the limitations in methodological quality, loss to follow-up and consistency of treatment strategy were scored. Disagreements were resolved by means of discussion, with arbitration by an independent trauma surgeon with a master in clinical epidemiology (N.S).
The search resulted in 488 potentially eligible studies, of which 17 full text articles were analysed for eligibility (Fig. 1). Eight studies met our inclusion criteria; however, two studies did not report sufficient data to be included in the meta-analysis leaving six studies available for meta-analysis [24–30]. These studies included a total of 135 patients in the union group and 230 in the nonunion group. Raw data were available for three out of the six studies [26,29,30]. Six patients (three in the union group and three in the
M.M.E. Wijffels et al. / Injury, Int. J. Care Injured 45 (2014) 1889–1895
1891
Idenficaon
References retrieved from data-bases PubMed, EMBASE, Web of Science, COCHRANE, CINAHL, Academic Search Premier, ScienceDirect, Springer, Wiley Blackwell,LWW, HighWire and Informahealth N=488 (aer removing duplicates)
References excluded Did not meet inclusion criteria. N=471 -19 language -41 case report -47 no abstract -16 chapters -3 comments Screening
-2 only abstract -2 animal studies -341 different subject
Eligibility
References Aer screening tles and abstracts N=17
Arcles excluded Did not meet inclusion criteria. N=9 -8 cohort studies
References
-1 no data on union-non union
Aer reading full-text arcle N=8 Arcles excluded No useful informaon N=2 -2 insufficient data
Included
References Aer reading full-text arcle N=6
Fig. 1. Study flow diagram of the systematic review.
nonunion group) had participated in both the studies of Buijze and Wijffels; their data were included only once in our meta-analysis. Description of included studies and outcome measure reporting The detailed description, reported outcomes, and methodological quality, based on GRADE guidelines, of the included studies in the meta-analysis are summarised in Tables 1 and 2. Analysis of raw data The combined raw data of the three studies [26,29,30] included the outcome data of 40 patients with union and 65 patients with
nonunion of the USP. No statistically significant differences between the combined union and nonunion groups were found for patientreported outcomes, range of motion, grip strength, and pain (Table 3). Meta-analysis: primary outcome Patient-reported outcome was measured in four studies [26,27,29,30] using (a combination of) different questionnaires: the Mayo wrist score [33] in three studies [26,27,30], the Gartland and Werley score [34] in two studies [26,30], and the Disabilities of the Arm, Shoulder and Hand (DASH) score [35] in three studies [26,27,29].
M.M.E. Wijffels et al. / Injury, Int. J. Care Injured 45 (2014) 1889–1895
1892 Table 1 Characteristics and results of the included studies. Study
n unions
n nonunions
Treatment of distal radius fracture
Mean age (range)
Analyzed outcome used in this meta-analysisb
Mean follow-up (range) (months)
Significant differences between groups
Buijze 2010a Kim 2011 Kraemer 2013
16 20 55
20 71 46
54 (25–75) 50 (18–88) 56 (18–81)
1,2,3,4,5,6,7 1,2,4,6,7 6,7
6 23 (12–48) 20 (6–49)
None None None
Mentzel 2001
6
29
59
6
12
Wijffels 2011a
16
18
Operative Operative Operative and conservative Operative and conservative Operative
56 (25–81)
2,3,4,5,6,7
31 (6–74)
22
46
Operative
64 (25–94)
1,4,5,6
14.9 (6–38)
Better range of motion in the union-group Better grip strength in the non-union group None
a
Zenke 2009 a b
Raw data available for analysis. (1) DASH, (2) Mayo Wrist Score, (3) Gartland and Werley score, (4) range of motion, (5) grip, (6) pain, (7) DRUJ instability.
Table 2 GRADE quality assessment of the included studies. Study a
Buijze 2010 Kim 2011 Kraemer 2013 Mentzel 2001 Wijffels 2011a Zenke 2009a a b c d e f g h i j k
Significant differences between groups
Study design
Inconsistencyb
Indirectnessc
Imprecisiond
Overall quality
None None None Better range of motion in the union group Better grip strength in the nonunion group None
Observational Observational Observational Observational Observational Observational
No No No No Seriouse No
No No Seriousj Seriousk Seriousi No
Seriousf No Seriousg Serioush Seriousf No
Low Low Very low Very low Very low Low
Raw data available for analysis. Inconsistency refers to unexplained heterogeneity of results. Indirectness refers to an indirect comparison of groups (e.g., loss to follow up and treatment strategy). Imprecision refers to studies with relatively few patients and few events and thus with wide confidence intervals around the estimate of the effect. No clear explanation for difference in grip strength. No DRUJ instability present. Only one questionnaire to include.
