Impact of Time to Fixation on Outcomes of Operative Treatment of Intra-articular Distal Radius Fractures

Abstract

Background:

The incidence of operative treatment of distal radius fractures (DRFs) has increased recently, but the optimal timing for surgical fixation remains unclear. We hypothesized that: (1) an increase in time to fixation of intra-articular DRFs would increase the likelihood of postoperative complications; and (2) increased time from injury to fixation would lead to longer surgical time and worse range of motion (ROM) outcomes.

Methods:

We retrospectively reviewed 299 fractures in 284 adult patients who underwent open reduction and internal fixation (ORIF) of a closed, intra-articular DRF at our institution over a 10.5-year period. Demographic information, time to surgery (TTS) from injury, surgical time, tourniquet time, complications, and final postoperative ROM were collected for logistic regression modeling to predict the risk of postoperative complication.

Results:

Twenty-seven (9.0%) patients experienced postoperative complications. The median TTS (Q1-Q3) for all patients was 7.0 (4.0-12.0) days. Patients who experienced an early postoperative complication had significantly longer median TTS (10.0 days) than those who did not (7.0 days). Patients with longer TTS were more likely to experience a complication (odds ratio, 1.11; 95% confidence interval, 1.04-1.19; P = .006). Tourniquet time and final wrist ROM were not related to TTS. A logistic regression analysis found that early complication rate doubles at 7.0 days after injury (from 3.5% to 6.9%).

Conclusions:

Patients with operative intra-articular distal radius fractures should ideally be fixed within 7 to 10 days of injury to minimize the risk of early postoperative complications. The tourniquet time and final ROM were not associated with time to surgery.

Introduction

Distal radius fractures are among the most common orthopedic injuries.^1,2 Despite a sizable body of literature available regarding distal radius fracture treatment, optimal management of these injuries remains controversial.^3,4 Indications, surgical approaches, fixation methods, and length of immobilization after surgery are well-studied variables affecting outcomes of distal radius fractures.^{3 -10} However, other potentially modifiable factors, such as time from injury to surgery, are less well studied. Several authors have investigated the relationship between timing of operative fixation of distal radius fractures and patient outcomes; however, a clear consensus regarding the ideal period of time within which surgical fixation should occur has not been established.^11,12 In the current literature, time has not been evaluated as a continuous variable, and the time points used to describe delayed fixation have been chosen by the investigating authors.

On the contrary, time from injury to surgery has been shown to affect outcomes in other upper extremity fractures. In the treatment of proximal humerus fractures, a delay of greater than 3 days has been associated with increased patient morbidity, and a delay of greater than 5 days has been associated with a significantly increased rate of complications.^13,14 Similarly, Chen et al ¹⁵ found significantly better Quick Disabilities of the Arm, Shoulder, and Hand (QuickDASH) scores, grip strength, and range of motion (ROM) in patients treated early versus treated late for operative ulnar styloid base fractures.

The purpose of our study was to determine whether an association exists between longer time to surgery and early postoperative complications in the treatment of intra-articular distal radius fractures. We hypothesized that an increase in time from injury to fixation of intra-articular distal radius fracture would be associated with an increased likelihood of postoperative complications. Second, we hypothesized that increased time from injury to fixation would lead to longer surgical time and worse ROM outcomes.

Materials and Methods

After approval by the institutional review board, we retrospectively reviewed the medical records of 484 nonconsecutive patients treated at our institution, including patients treated at both level 1 and level 2 trauma centers, from March 2007 to August 2017. A query of our institution’s online patient data support repository was performed to identify those patients who underwent open reduction and internal fixation (ORIF) of an intra-articular distal radius fracture, using the Current Procedural Terminology (CPT) codes 25607 (open treatment of extra-articular distal radial fracture or epiphyseal separation, with or without fracture of ulnar styloid, with or without internal or external fixation), 25608 (open treatment of intra-articular distal radial fracture or epiphyseal separation with internal fixation of 2 fragments), or 25609 (open treatment of intra-articular distal radial fracture or epiphyseal separation with internal fixation of 3 or more fragments). The CPT code 25607 was included as a quality control measure to allow for the identification of as many operatively treated distal radius fractures as possible. Those patients identified using code 25607 were subsequently reviewed by the study team to confirm that the fracture pattern was correctly coded as extra-articular and, if so, those patients were excluded. If patients coded with 25607 had in fact intra-articular fractures that were coded incorrectly, they were subsequently included in the study cohort.

All patients aged above 18 years, who underwent ORIF of a closed, intra-articular distal radius fracture by one of 4 fellowship-trained hand surgeons at our institution and its affiliate surgical centers with minimum 6-week clinical follow-up, were eligible for inclusion in this study. Those patients who underwent concomitant ORIF or closed reduction and percutaneous pinning (CRPP) of ulnar or radial shaft fractures, distal ulna fractures, carpal bone fractures, metacarpal fractures, and/or distal, middle, or proximal phalangeal fractures were excluded. In addition, any patient who underwent a concomitant nerve or soft tissue procedure, revision fixation of a distal radius nonunion or malunion, or carried the diagnosis of rheumatoid arthritis or similar condition compromising soft tissue was excluded. Patients treated in both outpatient and inpatient settings (including polytraumatized patients) were included in the study cohort.

Demographic information, such as patient date of birth, sex, smoking status (never, past history, or current), diabetes mellitus (presence or absence), and hand dominance, was collected for each patient. Information regarding each patient’s clinical care, such as date of injury, date of surgery, mechanism of injury (fall from height, fall onto outstretched hand, motor vehicle accident, sports injury, or other), fracture laterality, type of fixation (volar plate, dorsal bridge plate, dual plating, or other), surgical tourniquet time, and duration of follow-up, was also collected for each patient. Each patient’s medical record was reviewed to identify surgical complications, such as superficial infection, deep infection, delayed wound healing, tendon rupture, loss of fixation, loss of reduction, nerve or vascular injury, and fracture malunion (defined as fracture healing outside of the accepted standards of acceptable fracture alignment as outlined by the American Academy of Orthopaedic Surgeons at more than 3 months from the patient’s index surgical procedure). ⁴ Each patient’s initial injury radiographs were reviewed and assigned a corresponding fracture classification based on the AO (Arbeitsgemeinschaft für Osteosynthesefragen) fracture classification system. ¹⁶ The final ROM (flexion, extension, pronation, and supination) measured during physical examination by the patient’s hand surgery provider (attending surgeon, hand surgery fellow, or hand surgery advanced practice provider) was recorded from each patient’s final clinical follow-up visit.

Data are presented using counts and percentages for nonmissing categorical variables and means, standard deviations, medians, the 25th (Q1) and 75th (Q3) percentiles, or the ranges as appropriate for continuous variables. Statistical comparisons between patients with and without complications were conducted using a repeated-measures logistic regression model to adjust for intrasubject correlation when comparing 2 binary outcomes and results were presented using the odds ratio (OR) with 95% confidence interval (CI). Comparisons between continuous variables or among multiple categories were done using a generalized linear model with a log-normal distribution to account for intrasubject correlation as well as skewed data distributions.

Repeated-measures logistic regression models were also used to determine the cutoff values that best predicted a complication by maximizing the area under the receiver operating characteristic curve (ROC). The cut point was chosen to simultaneously maximize both sensitivity and specificity by applying Youden index to maximize the vertical distance of the ROC curve from the point (x, y) on the diagonal line (45° chance line). Unless otherwise stated, a P value less than .05 was considered statistically significant.

Results

A total of 299 fractures in 284 patients who met inclusion criteria were identified and reviewed (Table 1). Six patients were included after being identified using the quality control CPT 25607. The median age for the entire cohort was 55.7 years (Q1-Q3: 40.8-64.7). The median follow-up for the entire cohort was 29.0 weeks (Q1-Q3: 15.0-52.0). Eighty-nine patients had a minimum of 1 year of clinical follow-up, and 10 patients had a minimum follow-up of at least 2 years.

Table 1.

Demographics and ROM by Postoperative Complication Status.

	Total (N = 299)	Patients without complications (N = 272)	Patients with complications (N = 27)	OR, 95% CI	P value a
Age, y				0.99 (0.96-1.02)	.43
Median	55.7	55.9	54.6
Q1, Q3	40.8, 64.7	41.1, 64.9	38.9, 59.3
Female sex	197 (65.9%)	180 (66.2%)	17 (63.0%)	0.87 (0.34-2.14)	.74
Nonsmoker	255 (85.3%)	231 (84.9%)	24 (88.9%)	1.42 (0.36-5.55)	.59
Type 2 diabetes status	32 (10.7%)	29 (10.7%)	3 (11.1%)	0.96 (0.24-3.77)	.94
Time to surgery, d				1.11 (1.04-1.19)	.006
Median	7.0	7.0	10.0
Q1, Q3	4.0, 12.0	4.0, 11.5	6.0, 16.0
Tourniquet time, min				1.01 (1.00-1.03)	.073
N	290	264	26
Median	69.0	67.5	80.5
Q1, Q3	55.0, 90.0	55.0, 90.0	62.0, 107.0
Plate type					.44
Dual	9 (3.0%)	7 (2.6%)	2 (7.4%)	2.75 (0.44-17.08)
Volar	244 (81.6%)	221 (81.3%)	23 (85.2%)	Reference
Bridge	30 (10.0%)	29 (10.7%)	1 (3.7%)	0.33 (0.34-3.21)
Other	16 (5.4%)	15 (5.5%)	1 (3.7%)	0.64 (0.06-6.45)
AO classification					.33
B1	4 (1.3%)	4 (1.5%)	0 (0.0%)
B2	3 (1.0%)	3 (1.1%)	0 (0.0%)
B3 b	16 (5.4%)	13 (4.8%)	3 (11.1%)	1.07 (0.23-5.01)
C1	93 (31.1%)	89 (32.7%)	4 (14.8%)	0.32 (0.08- 1.22)
C2	102 (34.1%)	92 (33.8%)	10 (37.0%)	0.77 (0.27-2.18)
C3	81 (27.1%)	71 (26.1%)	10 (37.0%)	Reference
Flexion, °					.99
N	259	233	26	1.00 (0.98-1.02)
Median	60.0	60.0	63.0
Q1, Q3	45.0, 75.0	45.0, 75.0	50.0, 75.0
Extension, °				0.99 (0.97-1.01)	.15
N	263	236	27
Median	70.0	70.0	70.0
Q1, Q3	50.0, 80.0	51.0, 80.0	40.0, 80.0
Pronation, °				1.00 (0.95-1.06)	.87
N	222	198	24
Median	90.0	90.0	90.0
Q1, Q3	85.0, 90.0	85.0, 90.0	85.0, 90.0
Supination, °				1.00 (0.97-1.04)	.77
N	228	204	24
Median	90.0	90.0	90.0
Q1, Q3	80.0, 90.0	80.0, 90.0	80.0, 90.0
Total range of motion, °				1.00 (1.00-1.01)	.17
N	298	271	27
Median	280.0	280.0	280.0
Q1, Q3	155.0, 325.0	145.0, 325.0	230.0, 340.0

Note. ROM = range of motion; OR = odds ratio; CI = confidence interval; Q1, Q3 = the 25th (Q1) and 75th (Q3) percentiles; AO = Arbeitsgemeinschaft für Osteosynthesefragen.

^aP values were calculated using a repeated-measures logistic regression model.

^bB1, B2, and B3 combined for OR estimate and computation of P value due to small sample sizes.

Complication Rate

Early postoperative complications were observed in 27 fractures (9.0%) (Table 2). The patients who experienced a postoperative complication had a slightly lower median age (54.6 years, Q1-Q3: 38.9-59.3) than those who did not (55.9 years, Q1-Q3: 41.1-64.9), but the association was not statistically significant. No other significant differences in demographic variables were observed between fractures that did and did not lead to an early postoperative complication (Table 1).

Table 2.

Fractures With Early Postoperative Complications.

	Time to fixation (days)	AO fracture type	Complications
1	23	C3	Fracture malunion
2	22	C1	EPL tendon rupture
3	18	C2	Loose hardware requiring removal of hardware
4	17	C1	Median neuropathy at the wrist
5	16	C3	Deep infection
6	15	C3	FPL tendon rupture
7	14	C2	Fracture malunion
8	11	C3	FPL tendonitis from prominent hardware
9	11	C3	Fracture malunion
10	10	C3	Hardware failure (screw breakage) requiring removal
11	10	C2	Fracture malunion
12	9	C2	Radial artery laceration requiring repair
13	7	C2	FCR tendonitis
14	7	C3	Palmar cutaneous branch of median nerve (PCBMN) palsy
15	7	C2	Acute carpal tunnel syndrome
16	6	C3	First dorsal compartment tenosynovitis
17	5	C2	EPL tendon rupture
18	5	C2	Complex regional pain syndrome
19	5	C3	Extensor tenosynovitis requiring extensor tenolysis
20	3	C3	Deep infection
21	2	C2	EPL tendon rupture
22	32	C1	Superficial wound infection
23	28	C3	Superficial wound infection, wound dehiscence, or flexor tenosynovitis
24	15	B3	Superficial wound infection
25	10	C1	Superficial wound infection
26	4	B3	Superficial wound infection
27	1	B3	Superficial wound infection

Note. AO = Arbeitsgemeinschaft für Osteosynthesefragen; EPL = extensor pollicis longus; FPL = flexor pollicis longus; FCR = flexor carpi radialis; PCBMN = palmar cutaneous branch of median nerve.

Significantly, 14 patients elected to undergo subsequent removal of their locked volar plating construct owing to persistent pain and/or subjective wrist stiffness. These plate removals were not included in the postoperative complication group as this represented an elective removal of implants without objective plate-related complications.

Time to Surgery

Overall, the median time from injury to surgery was 7.0 days (Q1-Q3: 4.0-12.0). The median time to surgery was longer in patients who experienced an early postoperative complication (10.0 days, Q1-Q3: 6.0-16.0) compared with those who did not (7.0 days, Q1-Q3: 4.0-11.5). A repeated-measures logistic regression model determined that, as time to surgery increases, the likelihood of a postoperative complication significantly increases (OR 1.11; 95% CI, 1.04-1.19; P = .006). A cut point analysis found that dichotomizing at 10.9 days simultaneously maximized the sensitivity (59.3%) and specificity (75.0%) for predicting a complication; however, the model concordance index (c-index) was only moderate at 0.71 (95% CI, 0.60-0.82). Nevertheless, our model did find that the early complication rate doubles at 7.0 days after injury (from 3.5% to 6.9%) and triples at 11.3 days after injury (from 3.5% to 10.4%).

Tourniquet Time

The median tourniquet time across all fractures was 69.0 minutes (Q1-Q3: 55.0-90.0). A trend toward longer tourniquet time was observed in those patients who experienced a postoperative complication (80.5 minutes, Q1-Q3: 62.0-107.0) compared with those who did not (67.5 minutes, Q1-Q3: 55.0-90.0); however, a repeated-measures logistic regression model did not find a significant linear increase in postoperative complications as tourniquet time increased (OR 1.01; 95% CI, 1.00-1.03; P = .073).

Final Wrist ROM

No significant correlation between wrist flexion (P = .99), wrist extension (P = .15), pronation (P = .87), supination (P = .77) or total ROM (P = .17) at final follow-up were observed between fractures that did and did not lead to an early postoperative complication (Table 1). In addition, no significant correlations were observed between total ROM (P = .89) or tourniquet time (P = .69) and time to surgery.

AO Fracture Type

The most commonly observed AO fracture types were C2 (n = 102) and C1 (n = 93; Table 3). No significant differences in tourniquet time (P = .25), time to surgery (P = .91), or total ROM (P = .38) were observed based on AO fracture type.

Table 3.

Outcomes by AO Classification.

	B1 (N = 4)	B2 (N = 3)	B3 (N = 16)	C1 (N = 93)	C2 (N = 102)	C3 (N = 81)	P value a
Time to surgery, d							.91
N	4	3	16	93	102	81
Median	11	6	8.5	8	7	6
Q1, Q3	7.0, 14.5	5.0, 25.0	4.0, 11.5	5.0, 12.0	5.0, 11.0	4.0, 11.0
Tourniquet time, min							.25
N	4	3	16	91	99	77
Median	78	60	63	65	65.0	80
Q1, Q3	46.0, 114.5	36.0, 90.0	55.5, 87.5	55.0, 88.0	52.0, 90.0	64.0, 97.0
Total range of motion, °							.38
N	4	3	16	93	102	80
Median	337.5	360	335	305	270	267.5
Q1, Q3	312.5, 350.0	285.0, 360.0	122.5, 340.0	160.0, 330.0	150.0, 320.0	147.5, 320.0

Note. AO = Arbeitsgemeinschaft für Osteosynthesefragen; Q1, Q3 = the 25th (Q1) and 75th (Q3) percentiles.

^aGeneralized linear repeated-measures model with log-normal distribution used to compute P values.

Discussion

Greater time between injury and fixation of intra-articular distal radius fractures can lead to persistent pain, swelling, and localized inflammation, in theory increasing the risk for early postoperative complications. In this study, we found that patients who experienced early postoperative complications, such as superficial infection, deep infection, delayed wound healing, tendon rupture, loss of fixation, loss of reduction, nerve or vascular injury, and fracture malunion, had significantly longer times from injury to surgery. According to our logistic regression model, early complication rates were predicted to double at 7.0 days after injury and triple at 11.3 days after injury. In addition, we observed no association between mean tourniquet time and final postoperative wrist ROM in patients with longer time to surgery. The occurrence of a postoperative complication following intra-articular distal radius fracture is likely not attributable to time to fixation alone, and is likely multifactorial in nature; however, based on the observed study findings, greater time to fixation may contribute to the possibility of a postoperative complication occurring. We also observed no differences in complications rate based upon patient age, presence of diabetes, or smoking status. These findings are in contrast to prior literature regarding the impact of each of these patient factors on complication rates following fixation of distal radius fractures.^{17 -19}

The importance of timely follow-up and definitive fixation should be emphasized with patients at their initial injury encounter, particularly with regard to referring providers. Certain situations may warrant deferred surgical intervention, for example, in older adult or polytrauma patients who are poor candidates for surgery, or when the condition of the surrounding soft tissue has been compromised. Interventions aimed at reducing surrounding soft tissue swelling include appropriately fabricated splints and orthoses and elevating the affected extremity. Allowing soft tissue swelling to subside also provides a more supple soft tissue envelope that may decrease the difficulty of reduction and fixation.

Our study found an association between early postoperative complication rates and delayed time to surgery. These findings are similar to those of a randomized, multicenter trial by the Wrist and Radius Injury Surgical Trial (WRIST) group, which assigned patients with unstable distal radius fractures into 3 surgical interventions (external fixation, CRPP, and volar locking plate system).^1,20,21 These authors reported worse outcomes in the cohort with delayed surgical intervention, defined as greater than 7 days from injury. Specifically, they observed significantly increased digit stiffness (P = .03), chronic pain (OR, 3.65; 95% CI, 1.48-9.00), and forearm arc of motion but no associations with other postoperative complications. Several major postoperative complications included in our study, such as wound infection, tendonitis/tenosynovitis, and tendon rupture, were not collected in the WRIST study. We did not assess digit stiffness or patient-reported outcomes (PROs) to evaluate chronic pain, but the results appear to be consistent with previous literature.^11,12,22,23

The optimal time from injury to surgical intervention is poorly established, despite the number of recent studies investigating outcomes similar to ours. This may result from grouping patients by a variety of predetermined—and sometimes arbitrary—time cutoffs. As such, making direct comparisons between studies is imperfect. For example, Ahmad et al. retrospectively reviewed 190 patients who underwent surgical treatment (CRPP vs volar locking plate system) for distal radius fractures (extra-articular and intra-articular) and found no significant difference in complication rates between time interval groups (P = .17). The selection of these time intervals was designed to reflect the stages of acute inflammation, grouping patients by surgical intervention at 0 to 3 days, 4 to 10 days, and more than 10 days from injury. Several other studies similarly demonstrated no relationship between complication rate and time to surgery. Many do not comment on their definition of delayed time to surgery, which in some cases vary between 2 weeks and 21 days.^11,12,24 To our knowledge, our study is the first to examine time as a continuous variable. We observed increased rates of early postoperative complications in patients who were surgically treated after 7 to 10 days from injury. The discrepancies between our findings from these studies highlight the need to establish consistent, evidence-based guidelines between orthopedic societies.

Wrist ROM is an important functional outcome following surgical fixation of distal radius fractures. In this study, final wrist ROM was not significantly related to time to surgery. These results are generally consistent with recent literature, although 1 study found a 14° greater forearm arc of motion in patients with time to surgery less than 7 days.^{22 -25} Given time to surgery has not been shown to be associated with other functional outcomes based on PROs or return to work, it is unlikely that minor differences in wrist ROM, if present, have a clinically significant impact on patient quality of life. ¹¹

This study has several limitations. First, as a retrospective study, it is limited to data available in the electronic medical record and subject to weaknesses inherent to the study design. We did not have access to baseline or postoperative PROs, which may have been helpful in evaluating other valuable outcomes, such as functional changes and patient satisfaction. We did not conduct a power calculation before the current study given all available data from the past 10 years were included; however, a future study would only require 200 participants to have at least 80% power, assuming a similar complication rate of 9% and mean surgical delay times of 11.9 ± 7.8 and 8.1 ± 5.0 between those with and without complications, respectively. The use of a patient database as the primary source for patient identification may unintentionally omit patients treated at our institution who otherwise met study inclusion criteria but may have had either their procedure or demographic information incorrectly coded. In addition, the use of 6 weeks as the minimum postoperative follow-up in our inclusion criteria may not reflect the long-term outcomes. We observed an overall complication rate similar to other studies with relatively longer minimum follow-up times (ranging from 6 months to 2 years). As such, we feel our results are comparable to those obtained over longer follow-up intervals.^{1,11,12,20,21,22,25} Notably, whereas our cut point analysis found dichotomizing time to fixation at 10 days maximized sensitivity and specificity, the c-index associated with these findings was only moderate. It is possible that further investigation of long-term results may provide for a more predictive regression analysis model. The strengths of our study include the examination of time as a continuous variable as it relates to treatment outcomes.

Based on the results of this study, we recommend patients with intra-articular distal radius fractures deemed appropriate for surgical intervention ideally undergo operative fixation within 7 to 10 days of injury to minimize the risk of early postoperative complications. Patients who cannot undergo surgical fixation in this time period should be counseled that they may be at an increased risk of early postoperative complications.