Intramedullary nail fixation versus locking plate fixation for adults with a fracture of the distal tibia: the UK FixDT RCT

Internal pilot summary

The pilot took place in six centres over a period of 6 months. Screening logs were kept at each site to determine the number of patients assessed for eligibility and reasons for any exclusion. In addition, the number of eligible and recruited patients, and the number of patients who declined consent/withdrew, were recorded.

Main randomised controlled trial summary

All adult patients presenting at the trial centres with an acute fracture of the distal tibia were potentially eligible to take part in the trial. The broad eligibility criteria ensured that the results of the study could readily be generalised to the wider patient population. A computer-generated randomisation sequence, stratified by centre and age, was produced and administered independently by a secure web-based service, Warwick Clinical Trials Unit (WCTU). Randomisation was on a 1 : 1 basis to either IM nail fixation or locking plate fixation. Both of these operations are widely used within the NHS and all of the surgeons in the chosen centres were familiar with both techniques.

Baseline demographic data, radiographs and pre-injury functional data using the DRI and the OMAS questionnaire were collected. The patients were also asked to fill out the EuroQol-5 Dimensions (EQ-5D) health-related quality-of-life questionnaire twice at baseline; once to indicate their typical pre-injury health status and a second time to indicate their current post-injury status.

In conjunction with the clinical team, a research associate performed a clinical assessment and recorded any early complications at 6 weeks, and a radiograph was taken. A further clinical assessment and radiograph was also taken at 12 months postoperatively to detect late complications. Functional outcome, health-related quality of life and resource use questionnaires were collected at 3, 6 and 12 months postoperatively.

Inclusion criteria

The inclusion criteria for this trial were that the patient:

• was aged ≥ 16 years

• had any fracture that involves the distal tibial metaphysis, which was defined as a fracture extending within 2 Müller squares of the ankle joint18 (a Müller square is shown in Figure 1)

• had a closed fracture

• would, in the opinion of the attending surgeon, benefit from internal fixation of the fracture.

FIGURE 1.

Müller square. 18 Reproduced with permission. Copyright by AO Foundation, Biel, Switzerland.

Copyright by AO Foundation, Biel, Switzerland

Exclusion criteria

The exclusion criteria for this trial were that:

• there is, in the opinion of the attending surgeon, a contraindication to IM nailing – the presence of total knee replacement OR the medullary canal is too narrow OR there is a pre-injury deformity of the medullary canal OR it is not possible to achieve fixation of four cortices with screws distal to the fracture

• the fracture is open

• there is a contraindication to anaesthesia

• there is evidence that the patient would be unable to adhere to trial procedures or complete questionnaires

• the fracture extends into the ankle joint (i.e. intra-articular fracture).

Consent

Patients with a fracture of the distal tibia are admitted to hospital, through the emergency department, to a trauma ward. Patients are in pain and often treated with opiate-based painkillers during the initial treatment of their injury, so mental capacity may be impaired. However, although surgical treatment for a closed fracture of the tibia is urgent, it is not a time-critical intervention. In fact, traditionally, surgeons have deliberately waited a day or more before operating in order to let soft-tissue swelling settle. Once the leg is immobilised in a plaster cast, the pain is much better controlled, so mental capacity usually returns quite quickly following admission to the trauma ward.

During the internal pilot phase of the trial, we monitored the number of patients with impaired capacity and reviewed these data in conjunction with the Trial Steering Committee (TSC). In the context of this injury, by the time the patient was due to have surgery, none of the potentially eligible patients was judged to have reduced capacity according to the clinical team responsible for patient care. Therefore, informed consent from the patient was obtained by the local research associate before surgery. Patients were provided with verbal and written information about the study.

For those patients who withdraw from the trial after written consent had been obtained, data obtained up to the point of withdrawal have been included in the final analysis.

Sequence generation

A minimisation algorithm was used to allocate participants, so no random sequences were generated before the study. This is standard practice for trials conducted at WCTU.

Blinding

The type of fixation determines the site of clearly visible surgical scars. Specifically, the insertion of an IM tibial nail requires a surgical incision at the knee. Therefore, the patients could not be blind to their allocated treatment. In addition, the treating surgeons were also not blind to the treatment, but did not take part in the postoperative assessment of the patients. The functional outcome data were collected and entered into the trial central database via questionnaire administered by a research assistant/data clerk in the trial central office. The radiographs collected were reviewed by an independent assessor who was not involved in any other data collection or analysis.

Post-randomisation withdrawals

Participants could decline to continue to take part in the trial at any time and without prejudice. The participants were made aware that a decision to decline consent or withdraw would not affect the standard of care that they would receive.

Participants had two options for withdrawal:

1. Participants could withdraw from completing any further questionnaires but allow the trial team to continue to view and record any relevant hospital data that were recorded as part of normal standard of care, including radiographs and further surgery information.

2. Participants could withdrawal wholly from the study and only data obtained up to the point of withdrawal would be included in the final analysis of the study; thereafter, no further data were collected for that participant.

Once withdrawn, the patient continued under the care of their surgical team, as per normal clinical practice.

Intramedullary nail fixation

The IM nail is inserted at the proximal end of the tibia and passed down the hollow centre (medullary canal) of the bone in order to hold the fracture in the correct (anatomical) position. The reduction technique, the surgical approach, the type and size of the nail, the configuration of the proximal and distal interlocking screws and any supplementary device or technique were left at the discretion of the surgeon, as per standard clinical practice.

Locking plate fixation

A locking plate is inserted at the distal end of the tibia and passed under the skin onto the surface of the bone. Again, the details of the reduction technique, the surgical approach, the type and position of the plate, the number and configuration of fixed-angle screws and any supplementary device or technique were left to the discretion of the surgeon. The only stipulation was that fixed-angle screws must be used in at least some of the distal screw holes – this is standard practice with all distal tibia locking plates.

Risks and benefits

The risks associated with this study were predominantly those associated with the surgery: infection, bleeding and damage to the adjacent structures such as nerves, blood vessels and tendons. Participants in both groups had surgery and were potentially at risk from any/all of these complications. We believed that the overall risk profile was similar for the two interventions, but assessment of the number of complications in each group was a secondary objective of this trial.

Sample size

The minimum clinically important difference (MCID) is 8 points on the primary outcome (DRI) measurement scale. The DRI is a 12-question, patient-reported, functional outcome measure (physical exercise or sports, running, heavy physical work, heavy lifting, carrying a bag, leaning over a wash-stand, making a bed, moderate physical work, walks, mounting stairs, sitting still more than briefly and dressing or undressing) converted to a 100-point scale in which 0 points represents normal function and 100 points represents complete disability. At an individual patient level, a difference of 8 points represents the ability to climb stairs or run, with ‘some difficulty’ versus with ‘great difficulty’. At a population level, 8 points represents the difference between a ‘healthy patient’ and a ‘patient with a minor disability’. In addition, 8 points corresponds approximately to the clinically worthwhile benefit identified in other studies24 and is slightly lower than the 10-point difference between treatment group means in our pre-pilot study.

The SD of DRI score in our pilot study was approximately 20 points; the sample size was also estimated for a larger and smaller SDs to obtain an indication of the sensitivity to changes in this parameter. Assuming the distribution of DRI score in the study populations to be approximately normal, which is consistent with assumptions made for other reported trials using DRI as the primary outcome measure, Table 2 shows the total trial sample size with two-sided significance set at 5% for various scenarios of power and sample SD.

In Table 2, the number of 264 patients, represented in bold, represented the most likely scenario, based on our current knowledge, for 90% power to detect the selected MCID. Allowing a margin of 20% loss during follow-up, this gives a value of 320 patients in total. Therefore, 160 patients randomised to each group would provide 90% power to detect a difference of 8 points in the DRI score at 6 months, with 90% power at the 5% significance level.

Standard statistical summaries (e.g. medians and ranges or means and variances, dependent on the assumed distribution of the outcome) and graphical plots are presented for the primary outcome measure and all secondary outcome measures. Baseline data are summarised to check comparability between treatment arms and to highlight any characteristic differences between those individuals in the study, those who were ineligible and those who were eligible but withheld consent.

Analysis plan

The detailed statistical analysis plan was created by the trial statisticians and agreed by the DMC at the start of the study, in line with standard operating procedures at WCTU. Any subsequent amendments to this initial statistical analysis plan were clearly stated and justified to the DMC.

Software

All routine data reporting and final analysis was conducted using Stata® version 14 (StataCorp LP, College Station, TX, USA). 27

Data validation

Prior to formal analysis, data were checked for outliers, missing values and validated using the defined score ranges for all outcome measures. Queries were reported to the trial co-ordinator and investigated with the relevant recruiting centre if appropriate. Any subsequent changes to the database were recorded in accordance with the relevant WCTU standard operating practice and the UK Fixation of Distal Tibia Fractures (FixDT) trial data management plan.

Missing data

Data may not be available as a result of withdrawal of patients, lack of completion of individual data items or loss to follow-up. Reasons for data ‘missingness’ were ascertained and reported as far as possible. Any patterns of missingness were carefully considered, including, in particular, whether or not data could be treated as missing completely at random. No formal statistical testing was planned to assess missingness, but model assumptions were checked and patterns of missingness explored. If judged appropriate, missing data in the primary outcome (DRI score) was imputed using the ice (imputation by chain equation) procedure in Stata version 14. Any imputed data were on an individual-item level, as opposed to an overall score level. Reasons for ineligibility, non-compliance, withdrawal or other protocol violations are stated and any patterns are summarised.

Interim analyses

There were no planned interim analyses for the UK FixDT study.

Final statistical analyses

Exploratory analysis

To complement the preplanned analysis, a post hoc exploratory analysis using DRI score at all four time points was conducted. This analysis simplified longitudinal data collected at four time points to a single value, namely the area under the curve (AUC), and facilitated comparisons of the AUCs between treatment groups. It is advisable, in the presence of missing data, to use summary statistics generated by mixed models, as estimates will not be biased under the assumption that data are missing at random or are missing completely at random. 28 Therefore, we fitted a repeated measures mixed model, with the same fixed-effect structure as used in the primary analyses (adjusted for sex and age group), but with a three-level random-effects structure, in which observations (time points) were nested within participants and participants nested within recruitment site. The mixed and margins commands in Stata version 14 were used to obtain parameter estimates and standard errors (SEs) and the lincom command used to calculate AUC for each group. 27 AUCs associated with the two treatment groups were tested using a t-test.

Objectives

The economic evaluation was designed to estimate the costs of resource inputs and quality-adjusted life-year (QALY) profiles of each trial participant over the 12-month time horizon of the trial, allowing mean differences in costs and QALYs to be compared between the two arms of the trial. The primary objective was to evaluate the incremental cost-effectiveness of extra-articular distal tibia fractures treated with IM nail fixation versus locking plate fixation and to provide an estimate of the incremental net benefit.

Measurement of outcomes

Health-related quality of life was estimated using responses from the EQ-5D. This instrument facilitates the generation of a utility score from a person’s health-related quality of life. A health utility score refers to the preference that individuals have for any particular set of health outcomes. Patients completed the EQ-5D questionnaire at baseline and at 3, 6 and 12 months after randomisation, thus providing values of their health status at the time of questionnaire completion. Patients self-completed an additional EQ-5D at baseline, which assessed their pre-injury health-related quality of life. The standard UK (York A1) tariff values were applied to these responses at each time point to obtain utility scores. 29 The York A1 tariff set was derived from a survey of the UK general population (n = 3337), which used the time trade-off valuation method to estimate utility scores for a subset of 45 EQ-5D health states, with the remainder of the EQ-5D health states subsequently valued through the estimation of a multivariate model. 29 Resulting utility scores ranged from –0.59 to 1.0, with 0 representing death and 1.0 representing full health; values below 0 indicate health states worse than death. The second measurement component of the EQ-5D consists of a 20-cm vertical visual analogue scale (VAS) ranging from 100 (best imaginable health state) to 0 (worst imaginable health state), which provides an indication of the participant’s own assessment of their health status on the day of the survey. QALYs, which formed the main outcome of the economic analysis, were calculated using the AUC approach, assuming liner interpolation between the utility measurements. A five-level version of the EQ-5D (EuroQol-5 Dimensions, five-level version; EQ-5D-5L) was newly available at the start of the trial, but a UK tariff set for this version of the measure had not yet been published. Consequently, on advice from the TSC, it was decided to measure health-related quality-of-life outcomes using the EQ-5D.

Measurement of costs

Total costs for the two intervention arms were estimated by combining (1) resources used during the operation (implants and consumables), (2) total costs for the inpatient hospital stay and (3) patient-reported data on resource usage at 3, 6 and 12 months post randomisation. For the 3-month data, the recall period covered the period following hospital discharge; however, the recall period at subsequent time points covered the period since the last questionnaire was due to be completed. The trial participant questionnaires provided information on broader NHS and Personal Social Services (PSS) resource use as a result of the fracture. Specifically, the questionnaires captured the frequency of use of community-based health and social care services, number and duration of admissions to inpatient wards [classified as orthopaedics (your leg), orthopaedics (any bones), rehabilitation unit], number of diagnostic tests, use of outpatient services (classified as orthopaedics, physiotherapy, emergency department), medication use, and use of aids and adaptations. PSS included number of weeks of frozen/hot Meals on Wheels, laundry services and number of visits of carers and social workers. Total costs of resource usage were estimated by combining the unit cost data for each service with the resource usage data. In addition, the questionnaires captured private costs incurred by the patient as a result of the fracture, such as private physiotherapy. Productivity losses and lost income were also captured.

Cost-effectiveness analytical methods

An incremental cost-effectiveness analysis was performed, expressed in terms of incremental cost per QALY gained. Multiple imputation methods were used to impute missing data and avoid biases associated with complete-case analysis. The results of the economic evaluation were expressed in terms of incremental cost per QALY gained. A series of sensitivity analyses was undertaken to explore the implications of uncertainty on the incremental cost-effectiveness ratios (ICERs). Heterogeneity of cost-effectiveness results was addressed through the use of subgroup analysis. Further details are provided in Chapter 4.

Reporting

The reporting of the health economic evaluation is consistent with the Consolidated Health Economic Evaluation Reporting Standard statement. 30

Ethics committee approval

The UK FixDT trial was approved by the Coventry and Warwickshire REC on 6 November 2012 (REC reference number 12/WM/0340) and by the research and development department of each participating centre. The final, approved, study protocol has been published. 1

Trial Management Group

The day-to-day management of the trial was the responsibility of the trial co-ordinator, who was based at WCTU and supported by the administrative staff. This management was overseen by the Trial Management Group (TMG), which met monthly to assess progress. It was also the responsibility of the trial co-ordinator to undertake training of the research associates at each of the trial centres. The trial statistician and health economist were closely involved in the setting up of data-capture systems and the design of databases and clinical reporting forms.

Trial Steering Committee

A TSC was responsible for monitoring and supervising the progress of the UK FixDT trial. The TSC consisted of independent experts, a lay member and the chief investigator on behalf of the TMG. Membership of the TSC is given in Acknowledgements.

Data Monitoring Committee

The DMC was independent of the trial and was tasked with monitoring ethics, safety and data integrity. The trial statistician provided the data and analyses requested by the DMC at each of the meetings. Membership of the DMC is given in Acknowledgements.

Screening

Screening for potential participants began in April 2013. In total, 2118 patients with a tibial fracture were screened over 36 months. Screening data are presented to help assess the generalisability of the results to the overall population. Screening data are presented for all sites, with the exception of the Edinburgh site, which was closed to recruitment early in the main phase of the study.

There were significant differences in the total number of potential participants who were assessed between recruiting sites because of the varying practice at each site; for example, some sites screened children aged < 16 years but other sites were adult-only trauma centres. The wide variation in totals by centre indicates that some sites screened extensively and recorded any patient with a fracture involving the tibia; however, others appear to have only recorded fractures of the distal tibia on the screening logs.

The reasons why screened patients were not eligible (n = 1581) are detailed in Table 27, Appendix 1.

The most common reasons for participant ineligibility were that the fracture did not extend to within 2 Müller squares of the ankle joint (375/1581; 24%), the fracture was open (369/1581; 23%) and the fracture extended into the ankle joint (329/1581; 21%).

The total number of potentially eligible patients for the study was 537. The reasons why potentially eligible patients were not randomised (n = 216) are given in Table 28, Appendix 1.

Of the 537 eligible potential participants identified on screening logs, 40% (216/537) were not randomised. The conversion rate of eligible potential participants to randomised participants was therefore 60%. The most common reasons for non-participation were that there were no research staff available to consent the patient (n = 53; 25%) or that the surgeon had a preference to use an IM nail (n = 54; 25%). The other common reason was that the patient did not want to be part of a research study (n = 25; 12%). There were 18 potential participants for whom the reason for non-participation was ‘other’; these reasons included skin around ankle precluded plate fixation (n = 4), primary amputation (n = 3) and hind foot nail used (n = 2). The complete table of reasons why eligible potential participants were not randomised, by site, is given in Table 28, Appendix 1.

Data on sex and age were available for almost all of the patients screened (2003/2118; 95%). The largest subgroup of screened patients were men aged < 50 years, who accounted for 41% (819/2003) of the screened population.

Table 3 shows the age and sex of potentially eligible participants by randomised status (randomised or not randomised). The ages of both groups were similar and a t-test comparing the means of both groups showed no evidence of a difference in age (mean difference –0.9 years, 95% CI –4.1 to 2.3 years). Likewise, a two-sample test of proportions showed no evidence of a difference in sex distribution between the two groups (p = 0.818).

Recruitment

The trial planned to recruit 320 participants in total. Formal recruitment began in April 2013, with the first participant being randomised on 22 April 2013 at the lead site, University Hospitals Coventry and Warwickshire NHS Trust. Recruitment was undertaken at 28 sites over 36 months, with the final participant being randomised on 3 May 2016. For reporting purposes, the final participant has been included in the April 2016 recruitment month for ease of presentation. Details of site names and their opening dates are listed in Table 4.

The pilot phase was completed at the end of November 2013, with six sites open to recruitment. The planned recruitment rate was 0.75 participants pcpm. The recruitment rate averaged 0.6 participants pcpm during this phase, so the TSC recommended that the number of sites was increased from 18 to at least 24 to enable the trial to meet the target recruitment of 320 participants within the 30-month recruitment window. This resulted in a revised target of 0.6 participants pcpm.

Actual recruitment rates, by site, are shown in Table 29, Appendix 1. The overall observed recruitment rate for the main phase of the trial was 0.47 participants pcpm. This was somewhat lower than the revised target and so the number of trial sites was further increased to 28. One site was closed after just one patient was recruited, and screening data are not presented for this site.

Figure 2 illustrates the progression of recruitment over time, towards the required sample size target of 320. Recruitment slowed towards the end of the planned recruitment phase, so the recruitment window was extended, although this was still within the original timeline of the trial. Figure 3 shows the Consolidated Standards of Reporting Trials (CONSORT) flow diagram for the UK FixDT trial.

FIGURE 2.

Overall recruitment progression and target recruitment during the main phase (January 2014–April 2016) of the UK FixDT trial.

FIGURE 3.

The Consolidated Standards of Reporting Trials (CONSORT) flow diagram.

Table 5 describes the follow-up rates achieved during the trial. The expected follow-up rate used for sample size adjustment at 6 months was 80%. This was surpassed at the 6-month time point, with 88% of DRI assessments completed. Similar completion rates of 86% and 80% were achieved at the 3- and 12-month time points, respectively.

Tables 30 and 31, Appendix 1 demonstrate recruitment by stratification variables, namely age group and centre. There is good balance with respect to each factor, indicating that the minimisation procedure used to allocate treatment was implemented successfully.

The histograms in Figure 4 show the distribution of age in years, separately by sex. The overlaid kernel density estimator (the kdensity option of the histogram function in Stata version 14) shows the smoothed distribution and highlights the differences between sexes, with a noticeable but expected peak in the number of young male patients randomised. The mean difference in age between men and women was 6.8 years, 95% CI 3.2 to 10.4 years.

FIGURE 4.

Age distribution of randomised participants by sex. (a) Female; and (b) male.

Primary outcome

Figure 5a shows the trend in group DRI score means over the trial. Figure 5b shows the distribution of the DRI score at each time point, in which the middle bar is the median, the box represents the interquartile range and the whiskers extend to 1.5 times the interquartile range, with observations outside this range presented individually. Figure 5b demonstrates there is a treatment group difference in favour of the IM nail fixation group at 3 months, but this is reduced and not statistically significant at the 6-month time point and decreases further at the 12-month time point. Model assumptions were checked and appeared to be appropriate.

FIGURE 5.

The DRI. (a) Box plots of baseline and follow-up scores; and (b) overall trend in means and 95% CIs.

Table 11 shows the treatment estimates modelled using mixed-effects linear regression model, as previously described. The adjusted estimate of the treatment effect for the DRI score, at the 6-month post randomisation time point based on an intention-to-treat analysis, is 4.0 points (95% CI –1.0 to 9.0 points) in favour of the IM nail fixation group compared with the locking plate group. The p-value of 0.114 indicates that there is no evidence for a statistically significant difference in the DRI score between the two treatment groups at the 6-month post-randomisation time point. The prespecified MCID for the DRI is 8 points; therefore, at the 6-month time point we conclude that if there is a difference in the disability outcomes of the two groups, it is likely to be appropriately small as to be clinically unimportant. However, the 95% CI does include the prespecified MCID.

The adjusted estimate of the treatment effect at 3 months is 8.8 (95% CI 4.3 to 13.2) in favour of the IM nail fixation group compared with the locking plate group. The p-value of < 0.001 indicates that there is strong evidence for a statistically significant difference in treatment group means at 3 months. The estimated treatment effect is larger than the prespecified MCID for the DRI of 8 points, so this difference is likely to be clinically important to patients. At the 12-month time point, the p-value of 0.468 indicates that there is no evidence for a statistically significant difference in the DRI score between the two treatment groups at the 12-month post-randomisation time point, and the adjusted treatment difference is 1.9 points (95% CI 3.2 to 6.9 points) in favour of the IM nail fixation group compared with the locking plate group.

Secondary outcomes

Olerud–Molander Ankle Score questionnaire

The adjusted estimate of the treatment effect for the OMAS, at the 6-month post-randomisation time point based on an intention-to-treat analysis, is –6.0 points (95% CI –11.2 to –0.7 points) in favour of the IM nail fixation group compared with the locking plate group (Figure 6 and Table 12). The p-value of 0.026 indicates that there is evidence for a statistically significant difference in the OMAS between the two treatment groups at the 6-month time point.

FIGURE 6.

The OMAS questionnaire. (a) Box plots of baseline and follow-up scores; and (b) overall trend in means and 95% CIs.

TABLE 12 - Secondary patient-reported outcome measures: summary statistics, unadjusted and adjusted treatment effects at all time points

Mixed-effects regression model based on intention-to-treat analysis approach with allocated treatment group, age group, baseline pre-injury score and sex as fixed effects, and recruiting site as a random effect.

Outcome measure, time point	Treatment group				Difference (95% CI)		p-value
	IM nail fixation		Locking plate
	Mean (SD)	n	Mean (SD)	n	Raw	Adjusteda
OMAS (points)
3 months	42.3 (22.1)	130	36.0 (21.3)	139	–6.3	–7.0 (–12.0 to –2.0)	0.006
6 months	62.4 (23.1)	139	57.6 (24.9)	135	–4.9	–6.0 (–11.2 to –0.7)	0.026
12 months	73.8 (22.5)	20	70.8 (24.2)	29	–3.0	–3.6 (–9.1 to 1.9)	0.195
EQ-5D index score
Post injury	–0.003 (0.334)	158	–0.024 (0.311)	156	–0.021	–0.030 (–0.09 to 0.03)	0.331
3 months	0.546 (0.273)	134	0.499 (0.302)	142	–0.047	–0.058 (–0.12 to 0.00)	0.067
6 months	0.670 (0.265)	143	0.622 (0.275)	141	–0.048	–0.064 (–0.12 to –0.01)	0.029
12 months	0.722 (0.278)	128	0.731 (0.246)	130	0.009	–0.018 (–0.07 to 0.05)	0.525
EQ-5D VAS score
Post injury	46.6 (24.5)	158	45.3 (24.8)	157	–1.3	–0.8 (–5.7 to 4.0)	0.735
3 months	66.7 (20.5)	134	64.4 (21.1)	142	–2.3	–1.9 (–6.4 to 2.6)	0.418
6 months	75.0 (19.6)	143	71.6 (21.2)	141	–3.4	–2.5 (–6.8 to 1.8)	0.247
12 months	78.3 (20.5)	128	78.4 (20.8)	129	0.1	–0.2 (–4.6 to 4.2)	0.935

EuroQol-5 Dimensions index score

The adjusted estimate of the treatment effect for the EQ-5D index score, at the 6-month post-randomisation time point based on an intention-to-treat analysis, is –0.063 points (95% CI –0.12 to –0.01 points) in favour of the IM nail fixation group compared with the locking plate group (Figure 7 and see Table 12). The p-value of 0.033 indicates that there is evidence for a statistically significant difference in the EQ-5D score between the two treatment groups at the 6-month time point.

FIGURE 7.

The EQ-5D index. (a) Box plots of baseline and follow-up scores; and (b) overall trend in means and 95% CIs.

EuroQol-5 Dimensions visual analogue scale score

The adjusted estimate of the treatment effect for the EQ-5D VAS score, at the 6-month post-randomisation time point based on an intention-to-treat analysis, is –2.5 (95% CI –6.8 to 1.8) in favour of the IM nail fixation group compared with the locking plate group (Figure 8 and see Table 12). The p-value of 0.247 indicates that there is no evidence of a statistically significant difference in the EQ-5D VAS score between the two treatment groups at the 6-month time point.

FIGURE 8.

The EQ-5D VAS. (a) Box plots of EQ-5D VAS baseline and follow-up scores; and (b) overall trend in EQ-5D VAS means and 95% CIs.

Per-treatment analysis

Table 13 shows the results of per-treatment analyses, in which participants were analysed in per-treatment groups, that is, those who received IM nails compared with those who received locking plates. The five participants who did not receive either treatment were excluded from these analyses. An adjusted per-treatment analysis of the DRI scores at 6 months gave an adjusted treatment effect of 4.2 (95% CI –0.9 to 9.2) and a p-value equal to 0.103. Adjusted analysis at 3 and 12 months also gave similar results to the intention-to-treat analysis in Primary outcome.

Adjusted per-treatment analysis of the secondary outcome measures are also given in Table 13. All analyses yielded similar results to the equivalent intention-to-treat analysis.

Missing outcome data

The follow-up questionnaire completion rate was very good at all time points of the UK FixDT study. The expected loss to follow-up rate at the 6-month primary outcome time point was 20%; the actual rate was 12%. A summary of status at each follow-up time point is given in Table 5. In total, 284 out of 321 participants (88.4%) completed a DRI at the 6-month time point. The item-level missingness for the DRI score, and the secondary outcome measures, is given in Tables 32–34, Appendix 2. There were very few missing items in the DRI questionnaire at baseline (one item from 3816 items; < 0.1%), 3 months (1/3312; < 0.1%), 6 months (2/3408; < 0.1%) and 12 months (5/2640; 0.2%). Similar levels of missing data were seen across the OMAS questionnaire and the EQ-5D, with item-level missingness below 0.2% for all scales.

Table 14 shows the impact that including partial completion of the DRI using 11 or fewer items had on calculating the overall DRI score. There is very little difference in the adjusted treatment differences compared with the results given in Primary outcome based on fully completed questionnaires only. The largest differences are seen in the 12-month questionnaire data but this is because there are fewer observations and marginally more missing items than at 3- and 6-month data collection, resulting in a slightly larger treatment difference but being still in favour of IM nail fixation, as per the results in Primary outcome.

Subgroup analysis

At 6 months, the p-value of 0.516 indicates that there is no evidence for a significant interaction effect between age group and treatment group. The results of this analysis, and the subgroup analyses performed at 3 and 12 months, are given in Table 15.

Neither was there any evidence for a significant interaction effect between sex and treatment group (p-value of the interaction term = 0.384).

Exploratory analyses

Local complications related to the fracture or its treatment

Complications were assessed by the research team at the 6-week time point for almost all randomised participants (n = 314; 99% of participants who had a surgical procedure) and, again, in the patient reported questionnaires at 3, 6 and 12 months. Table 16 shows these complications by allocated treatment group.

The number of patients with symptoms related to infection was higher in the locking plate group (n = 32; 20%) than in the nail fixation group (n = 20; 13%), although this was not statistically significant (Fisher’s exact test p = 0.094). Twenty-one (13%) patients in the locking plate group and 14 (9%) in the IM nail fixation group were treated with antibiotics.

Other local complications were rare in both groups.

Data on ability/inability to bear weight on the injured leg were also collected at the 6-week visit in Table 16. In total, 33% of those allocated to nail fixation reported being fully weight-bearing at the 6-week visit, compared with only 14% of those allocated to locking plates. The difference in proportions is 0.19, 95% CI 0.10 to 0.41. The p-value for the difference is these proportions is p < 0.001, indicating strong evidence of a difference in the proportion of participants weight-bearing at 6 weeks.

Further surgery related to the fracture or its treatment is shown in Table 17. Where patients reported any further surgery, full details of the operation were requested and provided from the treating centre. There was more further surgery in the locking plate group (n = 19; 12%) than in the nail fixation group (n = 14; 9%), although this was not statistically significant (Fisher’s exact test p-value = 0.363).

Radiographic outcomes

An independent orthopaedic surgeon assessed the radiographic images, when available. The results of these independent assessments are given in Tables 18 and 19 for 6-week and 12-month images, respectively.

In addition to the independently assessed radiographs, the principal investigator at each recruiting site assessed key radiographic outcomes and reported these on the 6-week postoperative form. Comparisons between these site assessments and independent assessments were investigated, and there were good levels of agreement between assessors.

Systemic complications potentially related to the fracture or its treatment

Information regarding complications was also given through the SAE reporting pathway over the full 12 months of the trial. Data from related SAEs have been included in these complication profiles, with cross-referencing to ensure that appropriate counts are taken using multiple data sources. Not all of the venous thromboses or infections of the chest, urinary system, etc., will be related to the treatment but we have included all ‘potentially’ related events here for completeness. These data are summarised by treatment group in Tables 17 and 20.

Aim and perspective

The main objective of the health economic evaluation was to assess the cost-effectiveness of treating displaced extra-articular fractures of the distal tibia using IM nail fixation versus locking plate fixation. The primary analysis was undertaken from the perspective of the NHS and PSS, as recommended by the National Institute for Health and Care Excellence (NICE). 31 A societal perspective for costs was adopted for the sensitivity analysis and this included private costs incurred by trial participants and their families, as well as productivity losses and loss of earnings as a result of work absences.

The primary health economic outcome was the ICER attributable to IM nail fixation, which was calculated as the incremental cost per QALY gained 12 months after randomisation. No discounting of costs or health consequences was required as the trial-based economic evaluation was limited to a 12-month time horizon.

Measurement of resource use and costs

A comprehensive strategy was adopted to estimate the costs associated with distal tibia fixation using either locking plate or IM nail fixation. This included the (1) estimation of the initial fixation surgery costs and (2) estimation of broader health and personal social service resource inputs and broader societal resource inputs. All costs were expressed in pounds sterling and valued in 2014–15 prices. When appropriate, costs were inflated or deflated to 2014–15 prices using the Hospital and Community Health Services (HCHS) Pay and Price Inflation. 32

Costing of distal tibia fixation

The initial fixation surgery costs (intervention costs) were based on the initial hospital stay and associated operative costs, as reported in Table 22. The cost of distal tibia fracture fixation surgery was estimated using the NHS reference costs, specifically Healthcare Resource Group (HRG) code HT23D (major knee procedures for trauma). 33 According to this HRG code, operative costs for distal tibia fixation were £5315.47 for cases with a mean length of hospital stay of 5 days. Patient-specific costs for the initial operative period were identified using the average length of stay following primary surgery, as reported in the patient records. The mean length of hospital stay was 3.87 days for IM nail fixation versus 3.85 days for locking plate fixation and SEs were 0.34 and 0.33 days, respectively. The surgery cost of patients who stayed longer than 5 days was adjusted using the cost per excess bed-day figure of £327.00 for the same HRG code. For patients who stayed in hospitals < 5 days, we assumed that treatment costs were disproportionately weighted towards the first 3 days of each initial hospital admission. Thus, the cost to the NHS of a patient who stayed in hospital for 3 days was calculated as £5315.47 – (2 × £327), that is, the 5-day tariff minus the bed-day cost of £327 per each day not spent in hospital.

In addition, operative costs included the implants used during the surgery, namely nails, plates, locking screws/bolts, blocking screws and non-locking screws. Estimation of the number of each of these implants used involved prospectively recording the number of items used for each patient from the operation notes and radiographs. The total cost of implants for each patient was calculated by combining the resource inputs with their unit cost values; the unit cost values were derived from NHS trust finance departments.

Measuring broader resource use

Individual-level data on all significant health and personal social service and broader societal resource inputs were collected during the trial using follow-up questionnaires completed by trial participants. These data were collected at 3, 6 and 12 months post randomisation. The questionnaires captured the number and duration of admissions to inpatient hospital wards by ward type, number and type of diagnostic tests, use and type of outpatient hospital services, frequency of use and type of community-based health and social care services, medication use and aids and devices provided. In addition, the questionnaires captured the direct non-medical costs (including travel expenses) incurred by patients and their carers, as well as number of days off work and gross loss of earnings, attributable to the trial participant’s health state or contacts with care providers. Copies of the resource use questionnaires administered at each time point are provided in Appendix 3.

Valuation of resource use

The derivation of unit cost values was consistent with NICE’s Guide to the Methods of Technology Appraisal 201331 and included values extracted from the Department of Health and Social Care’s Reference Costs 2014–15,33 the Personal Social Services Research Unit (PSSRU)’s Unit Costs of Health and Social Care 2015,34 NHS Prescription Cost Analysis and the British National Formulary. 35 Table 35, Appendix 4, summarises the unit cost values and data sources for the broader resource use categories.

Further inpatient admissions following the initial operation were costed as minor knee procedures for non-trauma (HRG code HN25A) if the inpatient care involved procedures of the leg. However, if the inpatient admission was related to a surgical complication of the primary surgery, individual HRG codes that related to the subsequent operation procedures undertaken (e.g. debridement, metalwork removal, revision of internal fixation) were derived using the NHS HRG4 2014/15 Reference Cost Grouper software version RC1415 (NHS Digital, Leeds, UK). The Department of Health and Social Care’s Reference Costs 2014–1533 was used to assign the costs for each of the derived HRG codes. Subsequent inpatient care that was unrelated to procedures of the leg were also costed using Reference Costs 2014–1533 (Table 35, Appendix 4). The same approach was taken to cost subsequent hospital outpatient care, with resource use data on frequency of outpatient care being combined with the relevant unit costs.

Costs for community-based health services and PSS were calculated by applying unit costs extracted from national tariffs to resource volumes. Costs of medications for individual participants were estimated based on their reported doses and frequencies, when these were available, or based on an assumed daily dose using British National Formulary35 recommendations. When a dose range was reported as ‘as required’ or when the quantities were not recorded, we assumed a mean cost for that medication item based on the prescription cost analysis values (net ingredient cost per item). If the dose of the medication was missing, we assumed the patient received the same dosage as other trial participants who reported taking the same medication.

The costs of equipment that trial participants received to make their daily lives easier and manage their injury (aids and adaptations) were derived by combining the data on number and type of items received with their unit cost values. Unit costs were obtained from the NHS supply chain catalogue (https://my.supplychain.nhs.uk/catalogue; accessed 11 October 2017).

The costs of time taken off work were estimated by applying sex-specific median earnings data from the annual survey of hours and earnings for part-time and full-time work. 36 The employment status of trial participants was derived from self-reported work status information. Broader societal costs were calculated by combining the productivity losses and associated loss of earnings as a result of work absences and any privately incurred costs as a attributable to participants’ surgeries or impaired health states.

Calculation of utilities and quality-adjusted life-years

Participants’ health-related quality of life was assessed using the EQ-5D37 obtained at baseline and at 3, 6 and 12 months post randomisation. The EQ-5D defines health-related quality of life in terms of five dimensions: (1) mobility, (2) self-care, (3) usual activities, (4) pain/discomfort and (5) anxiety/depression. Responses in each dimension are divided into three ordinal levels coded: (1) no problems, (2) moderate problems and (3) extreme problems.

The EQ-5D health states are converted into a single summary index by applying a utility algorithm, which attaches values (weights) to each permutation of responses to the EQ-5D descriptive system. 37 EQ-5D preference weights have been elicited from general population samples in the UK using the time trade-off method. The resulting utility scores range from –0.594 to 1.0, with 0 representing death and 1.0 representing full health; values below 0 indicate health states worse than death. QALYs were calculated as the area under the baseline-adjusted utility curve and were calculated using linear interpolation between baseline and follow-up utility scores.

Missing data

Incomplete data are a particular issue in within-trial health economic evaluations and can result from item-level missingness; for example, when data for visit 2 are missing but data for visit 1 and all visits after visit 2 are available. 38 Consequently, a base-case analysis was constructed when missing data were imputed using fully conditional multiple imputation made chained equations, under the missing at random assumption. Multiple imputation under the missing at random assumption provides unbiased estimates of costs and health consequences. The missing at random assumption was tested through logistic regressions of missingness of costs and QALYs against baseline covariates.

Regression models were used to impute unobserved costs and QALYs at each time point and by treatment allocation using the baseline covariates (age, sex) as predictor variables. Costs and EQ-5D utility scores at each time point contributed as both predictors and imputed variables. The imputation was run 50 times, following the rule of thumb that the number of imputations should be similar to the percentage of incomplete cases. 39 The multiple imputation generated 50 data sets using predictive mean matching. Predictive mean matching provides plausible values when costs and QALYs are non-normally distributed. 40 In line with best practice, the MI model was validated by comparing the distributions of the imputed data with the observed data. 40

The multiply imputed data sets were analysed independently with bivariate regressions using a seemingly unrelated regression model (Sureg) to estimate the costs and QALYs in each treatment group over the 12-month trial horizon. Non-parametric bootstrapping was used to generate joint distributions of costs and outcomes from the original data set, and changes in costs and QALYs were calculated for each sample. A total of 1000 bootstrap samples were drawn and means for both incremental costs and incremental QALYs (with associated 95% CIs) were calculated.

The final step involved combining estimates from each imputed data set using Rubin’s rule to generate an overall mean estimate of costs and QALYs and the SEs. 40 The SE calculated through Rubin’s rules reflects the variability within and across imputations.

Analyses of resource use, costs and outcome data

Resource use items were summarised by trial allocation group and follow-up period and differences between groups were analysed using t-tests for continuous variables and Pearson chi-squared (χ²) test for categorical variables. Means and SEs for values of each cost category were estimated by treatment allocation and follow-up period. Statistical differences in mean costs by treatment allocation were assessed using Student t-tests. Mean total costs by treatment allocation and follow-up period were also estimated. Statistically significant differences in the mean total costs were assessed using non-parametric bootstrapping, based on 10,000 replications.

We calculated the proportion of patients reporting suboptimal health for each of the five dimensions of the EQ-5D. Patients were considered to be in suboptimal health if they reported moderate or extreme problems. We explored whether or not any statistical differences in suboptimal health-related quality of life existed between the two treatment arms at the different time points, using a Pearson chi-squared (χ²) test.

Cost-effectiveness analyses

Cost-effectiveness results are expressed in terms of the ICER and calculated as the difference between treatments in mean total costs divided by mean total QALYs. The bootstrap replicates from the non-parametric bootstrapping, described in Analyses of resource use, costs and outcome data, were used to populate cost-effectiveness scatterplots. Cost-effectiveness acceptability curves, which showed that the probability that IM nail fixation is cost-effective relative to locking plate fixation across a range of cost-effectiveness thresholds, were also generated based on the proportion of bootstrap replicates with positive incremental net benefits. The net monetary benefit (NMB) of using IM nail fixation versus locking plate fixation was also calculated across three cost-effectiveness thresholds, namely £15,000 per QALY, £20,000 per QALY and £30,000 per QALY. A positive incremental NMB indicates that the intervention is cost-effective compared with the alternative at the given cost-effectiveness threshold.

Sensitivity and subgroup analyses

Several sensitivity analyses were undertaken to assess the impact that uncertain parameters had on components of the economic evaluation. These involved re-estimating the main cost-effectiveness outcomes under the following scenarios: (1) restricting the analyses to complete cases (i.e. those with complete cost and outcome data over the 12-month follow-up period); (2) adopting a wider societal perspective that included private costs incurred by trial participants and their families, as well as productivity losses and loss of earnings owing to work absences; (3) estimating the cost-effectiveness under a per-treatment analysis; and (4) additionally adjusting the baseline analysis for pre-injury health-related quality of life, which was assessed using the EQ-5D at baseline.

Subgroup analyses were also conducted for the main cost-effectiveness results to explore heterogeneity in the trial population. These were conducted by (1) age group (< 50 and ≥ 50 years) and (2) sex (male and female).

Longer-term economic modelling

The study protocol allowed for decision-analytic modelling to extend the cost-effectiveness of IM nail fixation, drawing on best-available secondary data sources, supplemented when necessary by expert opinion.

Health-care resource use

Table 8 summarises the key resource inputs associated with the initial treatment of displaced extra-articular fractures of the distal tibia. Table 37, Appendix 4, presents details of broader health and social care resource use over the 12-month follow-up period for complete cases, disaggregated by resource category and period of follow-up. Generally, resource use at the aggregate level was higher for participants allocated to the locking plate than those allocated to IM nail fixation, but this was not always statistically significant. The exceptions were differences in mean total inpatient stay at 6 months (0 vs. 0.11 months), which was statistically significant at the 5% level (p-value of 0.03), and mean total outpatient care contacts at 6 months (3.64 vs. 4.78 mean outpatient contacts), which was also significant at the 5% level (p-value of 0.04). The reported number of days taken off work was generally higher for the locking plate arm. However, the mean difference (46.12 vs. 54.46) was significant at a 10% significance level at 3 months (p-value of 0.07). Regarding components of the resource categories, patients in the locking plate arm were more likely to use walking frames at 3 months (0.20 vs. 0.34), utilise more NHS physiotherapy (1.84 vs. 2.53) at 6 months and report higher use of ‘other’ medicines at 12 months (0.08 vs. 0.29).

Costs

Table 23 summarises the total costs associated with resource use during the trial period among complete cases by cost category and follow-up period. The mean intervention costs from admission until discharge were £5585 for IM nail fixation compared with £5615 for locking plate fixation; the mean difference of £30 was not statistically significant. The mean total NHS and PSS cost throughout the first 6 months post randomisation was £5876 for IM nail fixation and £6814 for locking plate fixation; the mean cost difference of £939 was statistically significant at the 5% level. The mean total NHS and PSS cost for the entire 12-month follow-up period was £6107 for IM nail fixation and £7102 for locking plate fixation; the mean cost difference of £995 was statistically significant at the 10% significance level. Total societal costs among cases with complete data are summarised in Table 24. The mean total societal costs throughout the first 6 months was £9793 for IM nail fixation compared with £12,178 for locking plate fixation; the mean cost difference of £2385 was statistically significant at the 10% significance level. The mean total societal costs are £9490 for IM nail fixation and £12,886 for locking plate fixation; the mean cost difference of £3396 was statistically significant at the 5% significance level.

Health outcomes

Table 25 details the number and proportion of individuals reporting each of the EQ-5D dimension levels at each time point. The proportion of trial participants reporting suboptimal health (moderate to extreme health outcomes) is also indicated for each dimension and the difference between the two treatment arms shown by p-values. With the exception of mobility at 3 months [IM nail fixation (81%) vs. locking plate fixation (89%)], which was statistically significant at the 10% significance level, there were no significant differences in the proportions of individuals reporting suboptimal health within dimensions between the two arms at each time point.

Cost-effectiveness results

The cost-effectiveness results are presented in Table 26 with locking plate fixation selected as the comparator. The analytic time horizon covers the entire 12-month follow-up period of the trial. Table 26 shows the joint distribution of incremental costs and outcomes for the base-case analysis, sensitivity analyses and subgroup analyses. The joint distribution of costs and outcomes for the base-case analysis, sensitivity analyses and subgroup analyses are graphically represented in Figure 9 using the NMB metric and in Figure 10 using scatterplots and cost-effectiveness acceptability curves. Plots for the remaining of the sensitivity analyses that are not shown in Figure 10 mirror the base-case analysis plot. As such, we elected not to present them.

FIGURE 9.

Forest plots of base-case, sensitivity and subgroup analyses showing the impact on incremental NMB at cost-effectiveness thresholds of (a) £15,000 per QALY, (b) £20,000 per QALY and (c) £30,000 per QALY.

FIGURE 10.

Cost-effectiveness scatterplots and acceptability curves at 12 months. Base-case analysis (NHS and personal social service perspective, imputed, intention-to-treat analysis): (a) scatterplot and (b) cost-effectiveness acceptability curve; complete-case analysis – (c) scatterplot and (d) cost-effectiveness acceptability curve; and imputed per-treatment analysis – (e) scatterplot and (f) cost-effectiveness acceptability curve.

Subgroup analyses

The subgroup analyses for age and sex revealed that there was significant interaction between treatment effect and age group, but this was significant at the 10% significance level (p-value of 0.09); the sex subgroup analysis did not reveal any statistically significant interactions. IM nail fixation lowered costs for patients aged < 50 years (–£1468, 95% CI –£3547 to –£291) and moderately increased QALY benefits (see Table 26). For patients aged ≥ 50 years, IM fixation reduced costs and moderately reduced QALY outcomes, with an ICER of £60,000 per QALY that fell within the south-west quadrant of the cost-effectiveness plane, that is, the average costs were less and average benefits were also lower for IM nail fixation than for locking plate fixation. The 95% CIs for both the costs and QALY estimates suggest considerable uncertainty surrounding the effects of IM nail fixation for this category of patients; this is graphically depicted in the NMB forest plot (see Figure 9). Cost-effectiveness acceptability curves for the subgroup analyses show the differences in probability of cost-effectiveness of IM nail fixation versus locking plate fixation (Figure 11). According to Figure 11, the probability that IM nail fixation is more cost-effective is higher for individuals aged < 50 years. There is no observable difference in probability of cost-effectiveness of IM nail fixation between male and female participants.

FIGURE 11.

Cost-effectiveness acceptability curves at 12 months (NHS and personal social service perspective, imputed). (a) Age; and (b) sex.

Eligible but did not participate

The reasons why eligible potential participants were not randomised (n = 216) are given in Table 28, Appendix 1. This information is key to determining the external validity of the trial, that is, the degree to which the result of the UK FixDT trial can be generalised to the broader population of patients with a fracture of the distal tibia.

Some of the reasons were inevitable and are unlikely to have created a selection bias within the trial. The NIHR Clinical Research Network is, generally, unable to provide routine weekend research associate support for clinical trials, so 53 out of these 216 patients (25%) were not recruited for the reason of ‘no research staff being available’. Similarly, there were 18 potential participants for whom the reason for non-participation was on obvious clinical grounds; these reasons included skin around ankle precluded locking plate fixation (n = 4), primary amputation (n = 3) and hind foot nails used (n = 2). Thirteen patients (13/216, 6%) declined any surgery, despite this being offered by the clinical team. The number of patients who declined the trial as they ‘did not want to be part of a research study’ (25/216, 12%) or ‘did not want to fill in a questionnaire’ (9/216, 4%) was small, which, once again, goes to show that patients with even severe injuries are keen to take part in research projects.

Patients who were not included in the trial because of a preference for one treatment or the other were of more potential concern. However, preference for a treatment did not seem to be a important issue for the patients; only 38 patients (38/216, 18%) expressed a preference, of whom 23 preferred a IM nail fixation and 15 a locking plate fixation. By contrast, there was a stronger treatment preference among the surgeons and 57 potentially eligible patients were excluded because the surgeon had a treatment preference for the patient. Of these 57 participants, 54 were because the surgeon had a preference to use an IM nail and three were because the surgeon had a preference to use a locking plate. The breakdown of patients who were potentially eligible but who did not participate by centre is informative. Some centres had no such patients; however, two centres had 13 (Royal Sussex County Hospital) and eight (Frenchay Hospital), suggesting that there was a ‘centre’ preference for IM nail fixation in those centres. The Royal Infirmary of Edinburgh demonstrated this phenomenon very early in their recruitment period and decided to withdraw from the trial completely after only one patient was recruited.

Overall, 60% of patients who were potentially eligible to take part in the trial were approached and agreed to take part. Only 17% of patients who were potentially eligible did not take part in the trial because of a preference for one treatment or the other; the majority as a result of surgeon preference rather than a patient preference. Table 3 showed that the baseline characteristics of those patients who were eligible and recruited versus those who were potentially eligible but not recruited were well matched.

Therefore, in summary, we can be confident that the results of this trial can be extrapolated to the wider population with a fracture of the distal tibia.

Treatment according to allocation

The CONSORT flow diagram (see Figure 3) shows that 156 out of the 161 patients allocated to IM nail fixation received the allocated treatment. Of the 160 patients allocated to locking plate fixation, 137 received the allocated treatment. Therefore, in total, 91% (293/321) of the patients who were randomised received their allocated treatment.

The fact that fewer patients crossed over from IM nail to locking plate fixation than from locking plate to IM nail fixation may, again, reflect a surgeon preference for IM nail fixation. However, the number of crossovers is too small to comment in any meaningful way. Importantly, the small number of crossovers overall makes it very unlikely that surgeon preference will have affected the result of the trial.

Recruitment by centre

The rate of recruitment by centres varied between < 0.1 pcpm (Basingstoke & North Hampshire Hospital) and > 1.4 pcpm (University Hospitals of Leicester). To a large degree this was determined by the size of the catchment area (population served) by the hospital. University Hospitals of Leicester has a very large catchment compared with Basingstoke & North Hampshire Hospital. Given that some centres were only open for a few months (Basingstoke & North Hampshire Hospital), and the number of patients presenting to each centre per month was relatively low, the variation in recruitment rate may also have been affected by simple variation in presentations over a short period of time.

However, some large-catchment hospitals were open for a relatively long time and still had a low rate of recruitment, for example Frenchay Hospital and University Hospitals Coventry & Warwickshire. Therefore, it is more likely that the rate of recruitment in those centres was affected by local recruitment issues. Several centres reported being affected by staff shortages (King’s College Hospital, Aintree University Hospital). Only one centre (Royal Infirmary of Edinburgh) withdrew because of lack of equipoise in the local surgeons.

Patients

The baseline demographic and clinical characteristics of both treatment groups were well balanced by the randomisation process (see Table 6). The baseline patient-reported outcome measures also showed good balance, with similar mean scores being seen across all of the outcomes measures.

Of the 321 patients randomised into the trial, 70% had no preference for which treatment they received. This is further evidence that the patients did appear to have equipoise having read the participant information sheet and discussed the trial with the clinical team, as well as with their friends and relatives.

The most common mechanism of injury was a ‘low-energy fall’ and the great majority of patients did not have any other injuries at the time of their fracture, that is, the majority of distal tibial fractures are isolated injuries rather than one of multiple injuries in the context of high-energy polytrauma (see Table 6).

Surgeons

Surgery for fracture of the distal tibia is potentially complicated and may be technically challenging, depending on the fracture pattern and associated soft-tissue injury. It is, therefore, unsurprising that the majority of operations in both groups were performed by consultant surgeons: 90 (56%) of IM nail fixations and 99 (62%) of locking plate fixations. The great majority of the other operations being performed by a trainee surgeon under supervision by a consultant. As surgery to fix the distal tibia is technically demanding, surgical assistants are often required. There was, on average, 2.5 surgeons present in the operating theatre for both the IM nail fixations and locking plate fixations.

Surgery

Full details of the surgery performed are given in Table 8. Of note, the duration of surgery was exactly the same in both groups of patients, with a mean duration of 2 hours and 4 minutes of operating time. This contrasts with some previous reports, in which IM nail fixation was found to be a quicker procedure. 14

The use of locking plate fixation provides more options for screw fixation distal to the fracture than IM nail fixation. This is inherent in the design of the implants and, indeed, is one of the theoretical advantages of locking plate fixation. As expected, the number of screws used in the distal part of the plates was greater than the number used in the nails. Five or six distal locked screws were used most commonly in the locking plate fixation group.

The use of blocking (Poller) screws to augment the IM nail fixation of metaphyseal fractures, such as those in the distal tibia, is the subject of much debate in the orthopaedic trauma community. However, surgeons in this trial only used blocking screws in 21% of fractures fixed with a tibial nail. The number of cases in which blocking screws were used is too small to comment on differences between centres and certainly too small to identify preferences between surgeons.

Intraoperative complications, including extension of the distal tibial fracture into the ankle joint, were rare: 2% in both groups. This is reassuring as one of the often quoted ‘risks’ associated with IM nail fixation is that the nail may propagate the fracture into the weight-bearing portion of the distal tibial articular surface at the ankle.

Table 9 shows that the number of patients who crossed over to the other treatment group after randomisation; this is inevitable in trials of surgical interventions and may be because of intraoperative clinical reasons or surgeon preference. Of course, there may be some overlap between these reasons, as an individual surgeon may feel more confident using one technique over another. In technically demanding surgery, such as fixation of fractures of the distal tibia, this may be a large effect. However, in the UK FixDT trial, over 91% (293/321) of participants received their allocated treatment. Therefore, we can be confident that the results of the trial were not unduly affected by crossover of treatment. Even so, we used the conservative intention-to-treat approach as the primary analysis.

Among those allocated to the IM nail fixation group, 97% (156/161) of participants received their allocated treatment, and in the locking plate group, 86% (137/160) of participants received their allocated treatment (there were three participant withdrawals before any intervention was undertaken: two were treated with external fixation and one was treated with manipulation under anaesthetic only). The small number of patients who crossed over were spread across all of the centres; the highest being eight patients (University Hospitals of Leicester), although this centre also had the highest overall recruitment (45 participants). No other centre had more than three patients crossover treatment, so there did not appear to be a centre-specific effect. A centre-specific effect in surgical trials may be because of the particular philosophy and/or training issues in that centre/regional training programme. Again, a preference for one technique may be more common in technically demanding surgery. The one centre (Royal Infirmary of Edinburgh) that withdrew from the trial after only one patient was recruited, cited a preference for tibial nail over locking plate and, therefore, a lack of equipoise. Surgeons at this centre have written a large number of the published literature on tibial nail fixation, so there may have been an unconscious bias towards this technique.

Follow-up rate

The sample size for the UK FixDT trial allowed for 20% loss to follow-up. We achieved > 80% follow-up at each time point. Therefore, we can be very confident that we have achieved sufficient statistical power according to our prespecified sample size calculation.

Primary outcome

The adjusted estimate of the treatment effect for the DRI score at 3 months, on an intention-to-treat basis, is 8.8 points (95% CI 4.3 to 13.2 points) in favour of the nail group compared with the locking plate group. The p-value of < 0.001 indicates that there is strong evidence for a statistically significant difference in treatment group means at 3 months. The estimated treatment effect was larger than the prespecified MCID for the DRI score of 8 points. Therefore, at the 3-month time point, the difference between treatment groups is likely to be clinically important to patients.

However, the adjusted estimate of the treatment effect at 6 months was only 4.0 (95% CI –1.0 to 9.0) in favour of the IM nail group compared with the locking plate group. Six months was the primary end point for the trial and is the point upon which the sample size calculation was based. The p-value of 0.114 indicates that there is no evidence for a statistically significant difference in the DRI score between the two treatment groups at 6 months. The 95% CIs suggest that if there is a difference in the DRI score, it is unlikely to be clinically important to patients. However, the upper 95% CI does include the prespecified MCID of 8 points.

At the 12-month time point, the p-value of 0.468 indicates that there is no evidence for a statistically significant difference in the DRI score between the two treatment groups, the adjusted treatment difference being 1.9 (95% CI –3.2 to 6.9) in favour of the IM nail fixation group compared with the locking plate group. At 12 months, the upper 95% CI excludes the MCID.

To complement the preplanned analysis, a post hoc exploratory analysis using DRI score at all four time points was conducted. This analysis simplified longitudinal data collected at four time points to a single value, namely the AUC, and facilitated comparisons of the AUCs between treatment groups. Using the predicted model parameter, the calculated AUC for those allocated to IM nail fixation was 350 units (SE 16.4 units) and for the locking plate group was 407 units (SE 16.3 units), yielding a between-group difference of 57 units (95% CI 12 to 103 units). Larger DRI scores represent higher levels of disability, so larger AUCs are also associated with increased levels of disability. A t-test comparing the values between groups yielded a p-value of 0.013, indicating there is a statistically significant difference in AUC between treatment groups. Clinically, this result may be interpreted as evidence that those allocated to locking plates experienced more disability over the 12-month follow-up period than those allocated to the IM nail fixation.

In designing the trial, we expected that any difference between these two treatments would most likely occur around 6 months. However, in the light of uncertainty about the time course of recovery, we also made assessments at 3 and 12 months. The results show that the largest difference in DRI score actually occurred at the earlier time point of 3 months. The difference in disability reduced at 6 months, and there was very little difference at 12 months. In retrospect, we could perhaps have used 3 months as the primary end point, although traditionally later time points have been used in all trials involving fractures. The very similar results at 12 months are not too surprising, as the great majority of fractures of the tibia would be expected to have achieved bony union at this time point; only very large differences in other aspects of recovery – for instance, muscle strength – would be demonstrated using any outcome measurement tool at this later time point. However, a minimum of 12-month follow-up is important in such trials so that the disability associated with delayed or non-union of the fracture can be assessed. Although rare in the context of this trial, non-union may be more much common in other fractures.

Overall, the primary outcome result of this trial is good news for patients, in that IM nail fixation is associated with faster recovery, that is, there is less disability during the early stages of recovery following a fracture of the distal tibia. This is also an important observation for the design of future trials in fracture healing.

However, it is important for patients to note that, although the DRI scores improve in both groups of patients over the 12 months following the fracture, they did not return to their pre-injury levels. The average DRI score at 12 months was still 23 points out of a maximum disability of 100 points.

To set these results into context, we repeated the systematic review of the evidence regarding this question at the end of the UK FixDT trial. As noted in Chapter 1, Background, there was little evidence in the literature regarding the pattern of recovery before the trial began. Our updated review found that four trials have reported since the UK FixDT trial. 41,44–47

Vallier et al. 44 randomised 104 patients with an extra-articular fracture of the distal tibia to either IM nail fixation or plate fixation. However, the plates used in this study were ‘non-locking’ plates, that is, traditional ‘large fragment medial plates’. The authors did not include any patient-reported outcome measures in the initial report of the trial. They noted a high rate of complications in both arms, including non-union and infection, which may have affected functional outcome but their results are confounded by the inclusion of patients with open fractures. The same group subsequently reported later functional outcomes from the participants in their trial using the Foot Function Index and Musculoskeletal Function Assessment questionnaires. 45 They found that ‘scores were not related to plate or nail fixation’ but that both Musculoskeletal Function Assessment and Foot Function Index scores ‘were worse when knee pain or ankle pain was present’.

Fang et al. 46 compared the results of external fixation combined with limited internal fixation with a plate, minimally invasive percutaneous plate fixation and IM nailing for distal tibia fractures. They concluded that ‘all achieved similar good functional results’. The authors did not report a disability rating, although there was no evidence of a difference between groups in the American Orthopaedic Foot and Ankle Society score. However, there were only 28 patients in each group in this trial.

Li et al. 41 compared external fixation with both plate and nail fixation. This trial concluded that all three methods were ‘efficient methods for treating distal tibia fractures’. There was no evidence of difference in the Mazur ankle score; no disability rating was reported. Again, this study was too small to draw firm conclusions, with only 137 patients distributed across all three groups.

Polat et al. 47 compared IM nail fixation with minimally invasive plate fixation. This trial did not find any difference in the Foot Function Index, but only 25 patients were included.

Therefore, there is little, if any, evidence with which to compare the main result of the UK FixDT trial. No other trial reported a disability rating. Other patient-reported scores were used in three trials, each showing no evidence of a difference between the treatment groups. However, each of these trials was considerably smaller than the UK FixDT trial and so subject to an increased risk of type II error, that is, these four trials may not have been large enough to detect a clinically important difference between IM nail fixation and locking plate fixation, even though one existed.

Secondary outcomes

Preplanned subgroup and secondary analyses

We preplanned a per-treatment analysis, expecting that some patients allocated to IM nail fixation would crossover to plate fixation, and vice versa. However, as the number of crossovers was actually small, we did not expect a difference between the intention-to-treat analysis and the per-treatment analysis. In keeping with this suggestion, the adjusted per-treatment analysis of the DRI scores at 6 months gave an adjusted treatment effect of 4.2 points (95% CI –0.9 to 9.2 points; p = 0.103). Adjusted analysis at 3 and 12 months also gave similar results to the intention-to-treat analysis.

We also preplanned a secondary analysis of the primary outcome measure, taking into account missing data. As there was very few missing data in the DRI forms at any time point, this analysis did not show any difference from the complete data analysis.

We also done preplanned subgroup analyses based upon sex and age, using < and ≥ 50 years as a surrogate for high-energy injuries in younger patients with normal bone density and lower-energy injuries in older patients with reduced bone density. At the primary outcome time point of 6 months, the p-value of 0.516 indicated there is no evidence for a significant interaction effect between age group and treatment group. Similarly, the p-value of 0.384 indicates there was no evidence for a significant interaction effect between sex and treatment group.

Health economic evaluation

The health economic evaluation is vital to the interpretation of the trial and not only for analysis of the initial cost of the implants and surgery. As the clinical data presented indicate that patients with a fracture of the distal tibia have persistent, measurable disability, even 1 year after their fracture, this injury is likely to have longer-term cost implications for both the health and social care systems.

Cost

The pre-pilot trial that informed the design of the UK FixDT trial suggested that there was considerable cost related to the choice of implant; at the time the cost of an IM nail and locking bolts was in the region of £350 and a distal tibia locking plate with screws was around £1200. Interestingly, in addition, operative costs included the implants used during the surgery, namely nails, plates, locking screws/bolts, blocking screws and non-locking screws. The estimation of the number of each of these implants used involved prospectively recording the number of items used for each patient from the operation notes and radiographs. The total cost of implants for each patient was calculated by combining the resource inputs with their unit cost values; the latter were derived from NHS trust finance departments.

Table 22 shows that the cost of the nail and four locking bolts had risen a little to approximately £450, whereas the cost of the medial distal tibial locking plate and eight locking screws had actually fallen to approximately £650. When the cost of disposable items (guidewire, reaming rod, etc.) is added to the cost of the IM nail, the difference between the costs of the two types of implant has greatly reduced in the last few years.

As a consequence, the trial found little difference in the cost of the initial surgical interventions. Neither was there much difference in the length of the index hospital stay. The total cost of the first admission to hospital was £5585 in the IM nail group and £5616 in the locking plate group.

However, between discharge from the index hospital stay and 6 months, both the inpatient and the outpatient health service costs were significantly greater in the locking plate group. The total NHS and PSS cost to 6 months was £5876 in the nail group and £6814 in the locking plate group, with a mean difference of £938, which was statistically significant (95% CI £84 to £1795; p = 0.04).

The difference in costs reflects the increased number of investigations, therapy visits and secondary interventions required in the locking plate group in the early phases of recovery. This fits with the clinical outcome data presented earlier, in that the number of patients having ‘problems’ was higher in the locking plate group than the IM nail fixation group. For example, patients with symptoms of wound infection would have more outpatient appointments, more blood and radiological tests and more in the way of antibiotic treatment than those without. Another factor contributing to the higher costs in the locking plate group may have been the higher proportion of patients who were not able to bear weight on their leg at 6 weeks. It is likely that this explains, at least in part, the increased number of therapy appointments in the locking plate fixation group.

Between 6 months and 12 months, there are, again, slightly higher costs in the locking plate group, but the difference between the groups was less than in the earlier phases of recovery. The mean cost difference to 12 months was £995.14, 95%CI £74.93 to £2069.63; p = 0.05. Further surgery, most commonly the removal of symptomatic metalwork, is likely to have happened later in the recovery phase when the surgeon and the patient were confident that the underlying fracture had healed. This may account for the increased costs in the locking plate group in the later phases of the trial.

A fracture of the distal tibia also has costs for the patient beyond the NHS and PSS perspective. The number of days off work in the first 3 months after injury was borderline statistically significant in favour of the IM nail fixation: 46 days in the IM nail group compared with 54 days in the locking plate group (p = 0.07). Depending upon the patient’s occupation, this may simply be a manifestation of the number of patients who were weight-bearing at this stage in their recovery, which was higher in the group having IM nail fixation. It may also reflect the greater number of days off work required to attend therapy, outpatient and inpatients visits in the locking plate group, or, most likely, a combination of the two.

Cost-effectiveness

As there is a small QALY gain in the nail group over the 12-month time frame of the trial, and the IM nail group incurred less cost, the cost-effectiveness evaluation is relatively straightforward to interpret; IM nail fixation dominates locking plate fixation. Assuming a cost-effectiveness threshold of £20,000 per QALY, there is a 97% probability that IM nail fixation is cost-effective.

The preplanned sensitivity analyses support this interpretation. The same result was noted with a complete-case analysis, an analysis including a societal perspective (i.e. including indirect costs, such as those associated with time off work) and an imputed attributable cost analysis. In the per-treatment analysis, there was a marginal QALY gain in the locking plate group, so IM nail fixation does not dominate in this scenario. However, the cost difference remained in the same direction and indicated that IM fixation was generally less costly than locking plate fixation.

In terms of the preplanned subgroup analyses, sex did not alter the interpretation; IM nail fixation dominates locking plate fixation in both men and women, that is, on average, it generates QALY gains and lower costs. The situation was not quite so clear in terms of age. For patients aged < 50 years, IM nail fixation again dominates, but there is much more uncertainty in patients aged ≥ 50 years. In this subgroup, although the costs of IM nail fixation were still lower, there were marginally fewer QALY gains, so the ICER was £60,000 per QALY gained in older patients.

Overall, the health economic evaluation provides strong evidence that IM nail fixation incurs less cost than locking plate fixation and marginally greater QALY gains. So, at any level of the cost-effectiveness threshold, IM nail fixation can be considered cost-effective.

Limitations

There are, of course, some limitations to the trial. Two hundred and sixteen potentially eligible participants were not randomised, which could pose a risk to the external validity (generalisability) of the trial. However, most of these patients were excluded for logistical reasons (e.g. no research staff available at the weekend), for good clinical reasons (e.g. declined any surgery) or because the patient ‘did not want to be’ part of a research study or fill out questionnaires’. There is no reason to suspect that these patients created a selection bias within the trial.

Patients who had a preference for one treatment or the other were of more concern. However, preference of treatment did not seem to be common among the patients; only 38 patients expressed a preference, of whom 23 preferred a IM nail and 15 a locking plate fixation. However, there was a stronger treatment preference among the surgeons, by whom 57 potentially eligible patients were excluded; 54 of these patients were excluded because the surgeon had a preference to use an IM nail. This does suggest that there was a preconceived preference for IM nail fixation, at least among surgeons in some centres. However, if patients deemed ‘suitable only for a nail’ were excluded, despite being eligible according to the trial criteria, this is only likely to have reduced the relative benefit of IM nail fixation compared with locking plate fixation. Despite this possible effect, the trial still showed an overall benefit of IM nail fixation.

Other possible limitations included post-randomisation crossover of patients from one group to the other. However, there were fewer than expected. Ninety-one per cent of patients received the treatment to which they had been allocated.

Similarly, some loss to follow-up was expected. However, the sample size was inflated to account for loss to follow-up of ≥ 20%, and < 20% of participants were lost from the primary outcome analysis at every time point.

Finally, as wound dressings after surgery are clearly visible, it was not possible to blind the patients to their treatment allocation.

Chief investigator and applicant

Professor Matthew Costa.

Trial co-applicants

Dr Juul Achten, Dr Nick Parsons, Professor Stavros Petrou, Professor Ian Pallister and Professor Sarah E Lamb.

Senior project manager

Susie Hennings.

Trial co-ordinators and data entry clerk

Nafisa Boota, Katie McGuiness, Ann-Marie Brennan and Jo Walter.

Trial statisticians

James Griffin and Dr Nick Parsons.

Health economists

Mandy Maredza, Professor Stavros Petrou and Melina Dritsaki.

Rehabilitation lead

Professor Sarah E Lamb.

Collaborators

Mr Alan Johnstone (Aberdeen Royal Infirmary); Mr Andrew Carrothers (Addenbrookes Hospital); Mr Mike McNicholas (Aintree University Hospital); Mr Nigel Rossiter (Basingstoke & North Hampshire Hospital); Mr Nikhil Kharwadkar and Miss Anna Chapman (Heartlands Hospital); Mr Tim White (Royal Infirmary of Edinburgh); Mr James Murray (Frenchay Hospital); Mr Mark Blyth (Glasgow Royal Infirmary); Mr Shivkumar Gopal (Hull Royal Infirmary); Mr Ian McMurtry (James Cook Hospital); Mr David Noyes (John Radcliffe Hospital); Mr Adel Tavakkolizadeh (King’s College Hospital); Professor Peter Giannoudis (Leeds Teaching Hospital); Mr Aradhyula Murty (North Tyneside General Hospital & Wansbeck General); Mr Ben Ollivere (Nottingham University Hospitals); Mr Mark Westwood (Derriford Hospital); Mr Mark Farrar (Poole Hospital); Miss Charlotte Lewis (Queen Alexandra Hospital); Mr Andrew McAndrew (Royal Berkshire Hospital); Mr Benedict Rogers (Royal Sussex County Hospital); Mr Andrew Gray and Mr Stephen Aldridge (Royal Victoria Infirmary); Mr Paul Dixon (Sunderland Royal Hospital); Miss Caroline Hing (St George’s Hospital); Mr John Kendrew (University Hospitals of Birmingham); Mr Jonathan Young (University Hospitals Coventry & Warwickshire); Mr Ashwin Kulkarni (University Hospitals of Leicester); Mr Gorav Datta (Southampton General Hospital); and Mr Kevin Smith (Royal Stoke University Hospital).

Research associates/fellows/nurses/assistants

Carol Carnegie, Simone Hargreaves, Melanie Harrison, Cara Kennelly, Safia Begum, Ewan Goudie, Steven Barnfield, Emma Sharp, Lisa Wilson, Alanna Milne, Louise Spoors, Sarah Brannigan, Bernadette Cook, Chris Herriot, Jessica Nightingale, Rosalyn Squire, Carrie Colvin, Sheeba Babu, Emily Bannister, Carrie Ridley, Karen Smith, Tracey Robson, Agne Sekmokaite, Lauren Day-Cooper, Louise Clarkson, Manjit Attwal, Maria Letts and Racquel Carpio.

Trial Steering Committee

Professor Matthew Costa, chief investigator.

Professor James Mason, chairperson.

Professor Simon Donell, independent member.

Mr Tim Chesser, independent member (subsequently chairperson).

Mrs Suzanne Jones-Griffith, independent lay member.

Mrs Ceri Jones, sponsor representative (observing).

Data Monitoring Committee

Dr Dawn Teare (chairperson).

Mr Richard Gibson.

Mrs Smitaa Patel.