Duration of Cast Immobilization in Distal Radial Fractures: A Systematic Review

Abstract

Objective  The duration of immobilization in distal radial fractures is disputed in the current literature. There are still no long-term superior outcomes of operative treatment in comparison to nonoperative treatment. A systematic review was initiated to assess the clinical controversy on the duration of the immobilization period for nonoperatively treated distal radial fractures.

Materials and Methods  A comprehensive search was performed in the PubMed, Embase, and Wiley/Cochrane Library databases and a manual reference check of the identified systematic reviews and meta-analyses was executed. Eligible studies were randomized controlled trials that compared two periods of immobilization, with reported functional, patient-reported, and radiological outcomes. Two reviewers independently agreed on eligibility, and assessed methodological quality and extracted outcome data.

Results  The initial search yielded 3.384 studies. Twelve trials, with 1063 patients, were included in this systematic review. Grip strength and patient-reported outcome were better in patients treated by a shorter period of immobilization. There was no difference in pain, range of motion, or radiological outcome between different periods of immobilization. Owing to heterogeneity of studies, data were unsuitable for pooling.

Conclusion  Included studies showed that there might be a preference for a shorter period of immobilization in nonoperatively treated distal radius fractures. Therefore, shortening the period of immobilization in distal radial fractures to a maximum of three weeks should be considered. Future research should include homogeneous groups of patients to draw valid conclusions on the appropriate period of immobilization for nonoperatively treated distal radial fractures.

Level of Evidence  This is a Level II study.

Systematic Review Registration Number  PROSPERO 2018 CRD42018085524.

Cast immobilization is the treatment of choice in the majority of patients with non- or minimally displaced distal radial fractures (DRF), resulting in satisfying outcome. ^{1 2} In most cases, a period of four to six weeks of cast immobilization is chosen. ^{3 4} However, the duration of cast immobilization is disputed in the current literature. Some authors believe that a period of three weeks of cast immobilization is sufficient, while others state that pain relief might be the only reason for cast immobilization. ^{5 6 7 8 9} The predominant risk of secondary displacement is the highest in the first two weeks after the injury and decreases to only 7–8% after this period. After six weeks, the risk of secondary displacement is negligible. ^{5 10 11} However, up to 30% of the patients suffer from long-term functional impairment after a nonoperatively treated DRF. It remains unclear why this is the case. ¹²

Furthermore, many patients with displaced DRF that were adequately reduced are treated nonoperatively with cast immobilization. Nonetheless, nonoperative treatment of reduced DRF competes with open reduction and internal fixation (ORIF) and the use of osteosynthesis has multiplied over the last years. ^{13 14 15} However, long-term outcomes between ORIF and nonoperative treatment are not different, especially not in non- or minimally displaced DRF. ^{16 17} Nonoperative treatment of DRF has a lower complication rate compared with ORIF with plate-osteosynthesis, 0–14 versus 17% respectively. ^{18 19 20 21} Kirschner-wiring has an even higher complication rate, 26–28%. ^{22 23} The question rises whether operative treatment should be preserved for those patients who are at risk of failure of nonoperative treatment. ^{21 24}

The present study was initiated to assess the clinical controversy on the duration of the immobilization period of nonoperatively treated DRF. Both non- and minimally displaced, and displaced and reduced DRF, were included. In the past, a Cochrane review on this subject has been performed and updated in the year 2008. ²⁵ Since then, no new systematic review has been published. Recently, two new studies have been performed comparing a regular period of immobilization to a period of immobilization of three weeks or less. ^{6 26} As recent publications show a trend toward a period of immobilization of three weeks or less for the treatment of DRF, a systematic review of prospective and randomized controlled trials (RCTs) was conducted. The purpose of this study was to determine whether the period of cast immobilization of DRF is related to functional outcome, patient-reported outcome, and treatment complications as secondary dislocation of the fracture and to investigate whether the period of immobilization of DRF could be safely shortened to three weeks.

Methods

A review protocol (PROSPERO 2018 CRD42018085524) was developed based on the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement ( www.prisma-statement.org ). A comprehensive search was performed in the bibliographic databases PubMed, Embase.com, and Wiley/Cochrane Library from inception up to 5th of April 2018, in collaboration with a medical librarian. The following terms were used, including synonyms and closely related words, as index terms or free-text words: “Conservative Treatment,” “Non-operative,” “Surgical Casts,” “Radius Fractures,” and “Adults.” The search was performed without date or language restriction. The full search strategies for all databases can be found in the Supplementary Material (online only). After deduplication, all titles were screened and appropriate abstracts were reviewed. A manual reference check of the identified systematic reviews and meta-analyses was performed. Furthermore, trial registers (EU Clinical Trails, clinicaltrials.gov, ISRCTN registry, Dutch Trial Register) were checked for unpublished articles.

Included Studies

In this systematic review, both prospective studies and RCTs were compared.

The regular period of immobilization was compared with a period of immobilization of three weeks or less. Different treatment modalities were divided into two groups: Group A: immobilization period of three weeks or less and Group B: regular immobilization period.

Data extraction on at least one of the following outcomes was requisite: functional outcome, patient-reported outcome, and radiological outcome. Studies that assessed operatively treated DRF, studies that involved fractures of the upper extremity other than DRF, studies that involved pediatric fractures, and studies that reported on radiological outcome alone were excluded.

Outcome Measures

Functional outcome was analyzed by measuring pain, grip strength, and range of motion. Pain after cast removal was measured by use of the Visual Analogue Scale (VAS). The VAS is a validated measurement tool for the assessment of chronic and acute pain. ^{27 28 29} On a continuous scale, patients record their subjective pain magnitude, with one end of the scale resembling being free of pain and the other end of the scale resembling the worst possible pain ever experienced.

Grip strength was expressed as a percentage of the contralateral wrist, measured with a dynamometer. Range of motion was measured by the use of a goniometer. The included studies reported the range of motion as a percentage of the contralateral wrist, degrees of range of motion, or sum of the degrees of range of motion.

Patient-reported outcome was analyzed by the use of the Patient Reporting Wrist Evaluation score (PRWE), Quick Disability of Arm, Shoulder and Hand score (qDASH), the Cooney-score, and the Gartland and Werley score. The PRWE, qDASH, and Cooney-score are patient-reported outcome questionnaires, resulting in a score between 0 and 100. For the PRWE and qDASH questionnaires, which only take subjective outcomes into account, 0 is the best possible outcome and 100 is the worst. ^{30 31} For the Cooney-score, which combines subjective and objective outcome, 0 is the worst possible outcome and 100 is the best. ³² The Gartland and Werley demerit score system for hand and wrist function is a combination of subjective and objective assessment of the range of motion, deformity, and complications after DRF. All points added together create a functional score, with excellent being 0 to 2, good 3 to 8, fair 9 to 20, and anything above 21 is labeled poor. ³³

Radiological outcome was analyzed by the use of the Lidström-score or by measuring dorsal angulation or radial length. The Lidström-score is a scoring system that reports the anatomical score. Deformity is determined by measuring the dorsal angle, amount of radial shortening, and loss of radial inclination. Distinction is based on the severity of deformation: severe, moderate, small, or insignificant. Severe deformity results in a poor Lidström-score, moderate deformity in a fair score, small deformity in a good score, and in case of insignificant deformity, an excellent score. ³⁴ The rate of dislocation was determined by degrees or millimeters of dorsal angulation and radial length. Initial dislocation in the included DRF was non- or minimal, or restored to non- or minimal after reduction. The initial dislocation and amount of dislocation after reduction were compared with the rate of dislocation after treatment.

Randomized controlled and prospective cohort trials were included if they compared the regular period of immobilization in the treatment of DRF to a period of immobilization of less than three weeks. Trials that included nonoperative treatment of both non- or displaced DRF were checked for eligibility. For trials to be eligible, a description of the duration of cast immobilization was a necessity. No restrictions on language of publication in used databases were imposed. However, retrieved articles in languages other than English, German, or Dutch, combined with the impossibility to translate the article to one of those languages, were excluded. Two authors (ED and TG) independently selected articles and extracted data, determined the risk of bias, with disagreements resolved by an independent judgement of a third author (FB). Data pertaining to quality of methods, participants, interventions, and outcomes were extracted. Methodological quality of trials was assessed with the internal and external validity criteria from the Cochrane checklist. Corresponding authors were contacted when there were specific questions to the design and outcomes of their studies.

Risk of Bias Assessment and Quality Scoring

A risk of bias assessment was performed according to the Cochrane Risk of Bias Tool. This tool reflects on selection bias, reporting bias, performance bias, detection bias, attrition bias, and bias due to reasons not covered in the other forms of bias. A judgement, based on various criteria, is made whether the article is at “Low” risk, “High” risk, or “Unclear” risk of bias. A study with low risk of bias was defined as fulfilling four or more of the criteria. Analysis of performance bias was disregarded in this systematic review, because blinding for period of immobilization was inapplicable. Attempt was made to contact the authors with the risk of bias assessment to give the opportunity to provide feedback. Nonetheless, no attempt was made to contact authors for publications before the year 2000.

Statistical Methods

Statistical analysis was performed by the use of Review Manager software, version 5.3, Cochrane Collaboration, London, UK. Homogeneous data were extracted by use of a fixed effect model, data were analyzed by use of forest plots. Protocol and study population were reviewed to determine clinical homogeneity. Statistical homogeneity was determined by use of the I ₂ test, with values less than 40% being considered to present no significant heterogeneity, visual inspection of forest plots and by use of the Q-test, with a significance level of p  < 0.05. When heterogeneity was present, possible sources of heterogeneity were explored, and if possible, the random effect model was used. Publication bias was examined by visual inspection of funnel plots produced by use of Review Manager software, version 5.3, Cochrane Collaboration, London, UK.

Results

The selection process of articles is shown in Fig. 1 . ³⁵ Five articles compared the results of cast immobilization to brace immobilization, for two to six weeks. ^{36 37 38 39 40} Unfortunately, these studies did not analyze outcome related to different immobilization periods and therefore these studies were excluded from this review. One study compared a period of immobilization of six weeks to a period of immobilization of four weeks in nondisplaced DRF; moreover, this study dates from the year 1973, and was excluded from this review. ⁴¹ All other studies compared the regular period of immobilization to a period of one to three weeks of cast immobilization and therefore fulfilled the inclusion criteria and were included in this systematic review ( Table 1 ).

Fig. 1.

Flow diagram of the selection process according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. ³⁵

Table 1. Articles eligible for systematic review.

	Author	Year	Article	n	Type of DRF	Mean FU	Study design		Risk of bias	Level of evidence
							Group A	Group B
1	Abbaszadegan et al 42	1989	RCT	80	ND	1 y	Bandage	Cast 4 wk	Unclear	II
2	Bentohami et al 6	2018	RCT	72	ND	1 y	Cast 3 wk	Cast 5 wk	Low	II
3	Christensen et al 7	1995	RCT	33	ND	9 mo	Cast 3 wk	Cast 5 wk	Unclear	II
4	Christersson et al 26	2018	RCT	109	DR	1 y	Cast 10 d	Cast 4 wk	Low	I
5	Davis and Buchanan 43	1987	RCT	55	ND	7 wk	Cast 1 wk	Cast 5 wk	Unclear	II
6	de Bruijn 48	1987	RCT	129	ND + DR	1 y	Cast 1 wk	Cast 4 wk	High	II
7	Dias et al 44	1987	PC	97	ND + DR	13 wk	Bandage	Cast 5 wk	Unclear	II
8	Jensen	1997	RCT	62	ND	6 mo	Cast 1 wk	Cast 3 wk	Unclear	II
9	McAuliffe et al 45	1987	PC	108	ND + DR	1 y	Cast 3 wk	cast 5 wk	High	II
10	Millet and Rushton 46	1995	RCT	90	ND + DR	3 y	Cast 3 wk	Cast 5 wk	Unclear	II
11	Stoffelen and Broos 47	1998	RCT	52	ND	1 y	Cast 1 wk	Cast 3 wk	Unclear	II
12	Vang Hansen et al 9	1998	RCT	100	ND + DR	1 y	Cast 3 wk	Cast 5 wk	Unclear	II
	Total			987

Abbreviations: DRF, distal radial fractures; DR, displaced & reduced; FU, follow-up; ND, non- or minimally displaced; PC, prospective cohort study; RCT, randomized controlled trial.

Group A = three weeks or less immobilization; Group B = regular period of immobilization.

In this review, we combined all patients who received the regular period of immobilization (Group B) and compared them to those with a period of immobilization of three weeks or less (Group A). In total, 987 patients (male/female, 112/875) were included in this systematic review; 431 of them were included in Group A.

Functional Outcome

Data for pain were available in six of the included studies ( Table 2 ). ^{6 26 42 43 44 45} Only one study reported significant difference, without describing mean values. ⁴⁵ Overall, the reviewed studies showed no difference in pain between Group A or B.

Table 2. Pain measured by VAS score.

Mean VAS after removal of the cast
	Author	Year	Type of DRF	Mean VAS after cast removal		p -Value
				Group A	Group B
1	Abbaszadegan et al 42	1989	ND	4	4.7	0.09
2	Bentohami et al 6	2018	ND	3.1	2.6	0.46
3	Christersson et al 26	2018	DR	2.5	2	0.41
4	Davis Buchanan 43	1987	ND	8.3	6.4	NS
5	Dias et al 44	1987	ND + DR	2.7 (ND) 2.8 (DR)	3.0 (ND) 3.5 (DR)	NS
						NS
6	McAuliffe et al 45	1987	ND + DR	–	+	0.004
Mean VAS after 1 year
1	Abbaszadegan et al 42	1989	ND	1.3	1.9	0.06
2	Christersson et al 26	2018	DR	0.2	0.1	0.91
3	McAuliffe et al 45	1987	ND + DR	+	–	0.02

Abbreviations: –, no known values; +, outcome in favor of this group; DRF, distal radius fractures; DR, displaced & reduced; ND, non- or minimally displaced; NS, not significant; VAS, Visual Analogue Scale.

Group A = three weeks or less immobilization; Group B = regular period of immobilization.

Grip strength was analyzed in five studies ( Table 3 ). ^{9 42 44 45 46} Overall, three of the five studies reporting on grip strength in non- or minimally displaced DRF reported a significant difference in favor of Group A. For displaced and reduced DRF, the result was not significant.

Table 3. Grip strength in % of uninjured side.

	Author	Year	Type of DRF	Mean FU	% Grip strength of uninjured wrist		p -Value
					Group A	Group B
1	Abbaszadegan et al 42	1989	ND	1 y	94%	78%	0.045
2	Dias et al 44	1987	ND + DR	3 mo (ND) 3 mo (DR)	76% 63%	58% 60%	0.000 0.540
3	McAuliffe et al 45	1987	ND + DR	1 y	−	−	0.001
4	Millet and Rushton 46	1995	ND + DR	6 mo	81%	65%	NS
5	Vang Hansen et al 9	1998	ND + DR	1 y	83%	90%	NS

Abbreviations: DRF, distal radius fractures; DR, displaced & reduced; FU, follow-up; ND, non- or minimally displaced; mo, months; NS, not significant; y, year.

Group A = three weeks or less immobilization; Group B = regular period of immobilization.

Range of motion was described in three studies ( Table 4 ). ^{26 42 46} Only one study reported a significant difference after one year in favor of Group A, and other studies did not find significant results.

Table 4. Range of motion after 1 year.

	Author	Year	Type of DRF	Range of motion	Range of motion		p -Value
					Group A	Group B
1	Abbaszadegan et al 42	1989	ND	Ext/Flex RD/UD	98% 98%	89% 90%	0.002 0.007
2	Christersson et al 26	2018	DR	Ext/Flex	15 degree	10 degree	NS
3	Millet and Rushton 46	1995	ND + DR	Sum	303.6 degree	290.6 degree	NS

Abbreviations: %, % of uninjured side; DRF, distal radius fractures; DR, displaced & reduced; Ext/flex: extension/flexion; ND, non- or minimally displaced; NS, not significant; RD/UD, Radial deviation, ulnar deviation; Sum, sum of degrees range of motion.

Group A = three weeks or less immobilization; Group B = regular period of immobilization.

Patient-Reported Outcome

Patient-reported functional outcomes were described in seven studies ( Table 5 ). ^{6 7 8 26 44 45 47} Only Bentohami et al. reported a significantly better patient-reported outcome in Group A for both PRWE and qDASH. ⁶ Other studies found no significant differences.

Table 5. Patient reported outcome.

	Author	Year	Type of DRF	PROM	Mean FU	Outcome		p -Value
						Group A	Group B
1	Bentohami et al 6	2018	ND	PRWE	1y	5	8.8	0.045
				qDASH	1y	0	12.5	0.026
2	Christensen et al 7	1995	ND	G&W-s	9m	1	1	NS
3	Christersson et al 26	2018	DR	G&W-s	1y	4.8	4.3	0.38
4	Dias et al 44	1987	ND + DR	G&W-% (ND) G&W-% (DR)	3m	24% (ND) 16.3% (DR)	6.4% (ND) 6.4% (DR)	–
5	Jensen 8	1997	ND	G&W-%	6m	68%	88%	NS
6	McAuliffe et al 45	1987	ND + DR	G&W-%	1y	85%	77%	NS
7	Stoffelen and Broos 47	1998	ND	Cooney	1y	86.8	82.2	NS

Abbreviations: Cooney, Cooney-score (100: best, 0: worst); DRF, distal radial fractures; DR, displaced & reduced; FU, follow-up; G&W-%, percentage of patients with excellent function; G&W-s, Gartland and Werley score, (0–2: excellent, 3–8 good, 9–20 fair, >21: poor); ND, non- or minimally displaced; NS, not significant; PROM, patient reported outcome measure; PRWE, Patient reported wrist evaluation (0: best, 100: worst); qDASH, Quick Disability of Arm, Shoulder and Hand (0: best, 100: worst).

Group A = three weeks or less immobilization; Group B = regular period of immobilization.

Radiological Outcome

Radiological outcome was assessed in nine studies, of which four reported on the Lidström-criteria ( Table 6 ). ^{6 7 8 9 42 44 45 46 48} Only one study reported a significant difference of excellent and fair outcome in favor of Group B. ⁴² Displaced and reduced DRF had worse outcomes compared with non- or minimally displaced DRF. In all studies, results were closely related.

Table 6. Radiological outcome.

	Author	Year	Type of DRF	Outcome	Radiological outcome		p -Value
					Group A	Group B
1	Abbaszadegan et al 42	1989	ND	L-ex L-f	68% 6%	91% 0%	<0.05 <0.05
2	Bentohami et al 6	2018	ND	L-ex L-f	97% 0%	97% 0%	NS NS
3	Christensen et al 7	1995	ND	DA RL mm	3.1 1.5	4.4 1.2	NS NS
4	de Bruijn 48	1987	ND + DR	RL mm	0.7	1.9	0.3
5	Dias et al 44	1987	ND + DR	L-ex (ND) L-f (ND) L-ex (DR) L-f (DR)	22% 28% 7.0% 42%	4% 19% 9% 38%	– – – –
6	Jensen 8	1997	ND	L-ex	100%	88%	NS
7	McAuliffe et al 45	1987	ND + DR	DA RL mm	8.5 3.3	8.9 3.8	NS NS
8	Millet and Rushton 46	1995	ND + DR	DA mm RL mm	4.6 2.8	6.2 3.4	NS NS
9	Vang Hansen et al 9	1998	ND + DR	DA degrees RL mm	4 9	5 9	NS NS

Abbreviations: DA, dorsal angulation in degrees; DRF, distal radial fractures; DR, displaced & reduced; L-ex, Lidström, excellent; L-f, Lidström, fair; ND, non- or minimally displaced; NS, not significant; RL mm, radial shortening in mm.

Group A = three weeks or less immobilization; Group B = regular period of immobilization.

Risk of Bias within Studies

Using the Cochrane Risk of Bias Tool, two studies were qualified as low-risk studies, two as high-risk studies, and in eight of the studies the risk of bias was unclear ( Table 1 and Fig. 2 ).

Fig. 2.

Risk of bias assessment.

Discussion

This systematic review assessed the clinical controversy on the duration of the immobilization period for nonoperatively treated DRF. Most DRF are treated nonoperatively by cast immobilization. Indications for nonoperative treatment vary among non- or minimally displaced DRF, displaced and adequately reduced DRF, and displaced DRF in patients who are not fit to undergo surgery. It is, therefore, important to investigate the different options of nonoperative treatment. The purpose of this study was to perform a systematic review and meta-analysis of the available publications on the duration of cast immobilization in DRF.

In 2003, Handoll and Madhok published the Cochrane review “Conservative interventions for treating DRF in adults.” ²⁵ While the review was edited in 2008, no changes to the conclusion were made since no updates or new systematic reviews were performed.

Until 2018, no new RCTs comparing the regular period of immobilization to a shorter period have been executed. In 2003, two RCTs were performed; however, both studies compared nonoperative treatment by cast immobilization to treatment by brace or splint, and did not compare different periods of immobilization. ^{39 40}

Two new studies have been conducted recently, comparing the regular period of immobilization to a period of three weeks of immobilization. ^{6 26} Both studies analyzed homogeneous patient populations and clearly reported their data with the use of standard deviations or interquartile ranges; nevertheless, they both analyzed different outcome measures.

The first study analyzed pain, Lidström-score, and also patient-reported outcome, using the PRWE and qDASH score, while the other analyzed grip strength, range of motion, pain, and patient-reported outcome using the Gartland and Werley score. ^{6 26} Consequently, these data could not be pooled to perform a meta-analysis. Quantitative analysis by meta-analysis was therefore abandoned and a systematic review, comparing the old and new data, was performed.

In this study, no difference in pain between Group A and B was found. However, one study did report a significant difference in favor of the short immobilization group, although mean values were not described. ⁴⁵ Therefore, these results were hard to interpret and clear conclusions could not be drawn.

Studies showed a possible benefit of shorter immobilization in grip strength, especially in the patients who suffered non- or minimally displaced DRF. In the three studies analyzing range of motion, three different methods were used. Again, results were too different to be pooled. However, as visualized in Table 4 , there seems to be a trend in range of motion, as the overall outcome of patients who were treated by a shorter period of immobilization was better, compared with the other group. However, comparing the range of motion to the contralateral side could be inaccurate; the reliability of this outcome can therefore be questioned.

Patient-reported outcome was analyzed by four different patient-reported outcome measures. Only the PRWE and qDASH are validated questionnaires, resulting in stronger evidence on hand and wrist function. ^{49 50} Only one study used these questionnaires, and showed a significant difference in favor of Group A. ⁶ Despite the fact that most studies did not use a validated tool to assess the patient-reported outcome, this study might have found important results, indicating the benefit of a period of immobilization of less than four weeks in non- or minimally displaced DRF.

Radiological outcome was measured by the Lidström-classification, dorsal angulation in degrees or millimeters, and radial length in millimeters. The Lidström-score, used in many articles of this review, might be less reliable than mean degrees or millimeters dislocation. Only one study reported an excellent outcome on the Lidström-score more frequently in patients who received the regular period of immobilization, although this study only analyzed non- or minimally displaced DRF. ⁴² Why only 68% of the patients who were treated in group A were scoring excellent on radial outcome after 1 year is unclear. Thirty-two percent of the patients scored a Lidström-score that was not excellent and had at least some secondary dislocation. Non- or minimally displaced DRF tend to dislocate less often than displaced DRF. Unfortunately, this outcome was not compared with patient-reported outcome; therefore, the clinical importance could be questioned.

It is believed that the purpose of cast immobilization, especially in nondisplaced DRF, is for pain relief, and not to prevent dislocation. ^{8 9 11 42} Therefore, it is unclear why the percentage found in this study is so high.

Overall, this systematic review of 12 included studies revealed that patients with DRF treated by a period of immobilization of 3 weeks or less had a functional outcome that was noninferior to the regular period of immobilization.

Results were limited by the results of bias and strength of the available evidence. Also, no distinction could be made between nondisplaced and displaced DRF. Taking into account that most studies were published before the year 2000 and also the most recent studies were included in this systematic review, published results are a good representation of the current evidence on this subject.

Consequently, more research is necessary to draw clear conclusions about the best period of immobilization in DRF. Future studies should use homogeneous groups of patients and report standard deviations to allow quantitative analysis to compare non- or minimally displaced and displaced and reduced DRF.

In conclusion, this study showed noninferiority of a period of immobilization of three weeks or less. A period of immobilization of three weeks or less might even be associated with better functional outcome. Therefore, shortening the period of immobilization in DRF should be considered.

Funding Statement

Funding None.