Background
Trauma to the upper extremity is common and such patients account for 10-30% of the attendance to an Accident and Emergency department [
1,
2]. Most patients are young men, and the majority of hand injuries are minor or moderate [
1‐
3]. However, hand injuries do have an impact on the patients` work capacity, daily activities and leisure activities [
4]. Thereby, the injuries have a profound power on the individual subjects as well as on the society. Irrespective of the variety of severity of hand and arm injuries, substantial costs, consisting of health care costs and costs of lost production, are induced to the society [
5,
6]. Even if the costs of hand injuries have been the focus in previous research, the more severe ones, some of them caused by machines [
6,
7], may induce substantial costs due to an increased time off work and to the subsequent permanent impaired hand and arm function [
8]. Our aim was to evaluate total costs and factors associated with such costs and to measure physical function (i.e. Disability of Arm, Shoulder and Hand; DASH) and quality of life (i.e. EQ-5D) in consecutively and randomly included patients for serious hand injuries.
Patients and methods
Patient material
Between February 2005 and April 2007, 153 patients with a major or severe injury according to the Hand Injury Severity Score (HISS) [
9] or the modified HISS [
10] were treated at our department. Totally 132 fulfilled the inclusion criteria, i.e. being able to communicate in Swedish and aged between 16–65, with a major hand injury, based on Hand Injury Severity Score (HISS) with scores >100 and with a severe hand injury with scores 50–100. We randomly, based on the individual physician on call´s decision and attention, asked 54 (45%) patients to participate in a prospective study. The patients were included independently of which weekday or month they were injured (results not shown). Nine patients declined participation in the study for various reasons. The remaining 45 (34%) patients were followed for 12 months after the injury (Table
1).
Table 1
Patient characteristics of included and not included patients with a severe or major hand injury
Gender
| | |
Female/Male | 9 (20)/36 (80) | 12 (14)/75 (86) |
Age
| | |
(years) | 42 (16–64) | 37 (16–65) |
HISS
| 154 (52–414) | 82 (51–268) |
Injury severity group
| | |
Major | 36 (80) | 29 (34) |
Severe | 9 (20) | 57 (66) |
Emergency surgery
| | |
(minutes) | 336 (32–1879) | |
Type of injury
| | |
Amputation | 13 (29) | |
Complex | 18 (40) | |
Nerve/Full-house | 13 (29) | |
Burn | 1 (2) | |
Reconstruction of blood vessels by microsurgery
| | |
yes | 21 (47) | |
no | 24 (53) | |
Cause of injury
| | |
Crush | 17 (38) | |
Cut | 27 (60) | |
Burn | 1 (2) | |
Type of work
| | |
Manual | 28 (62) | |
Non-manual | 11 (24) | |
Self employed | 2 (5) | |
Students/early retirement/long-term sick leave | 4 (9) | |
Methods
Demographic data were collected from the medical notes of all 132 patients. The severity of the injury was classified with HISS or with the modified HISS when applicable. At three, six and 12 months the 45 included patients answered the region-specific Disability of the Arm, Shoulder and Hand (DASH) outcome measure [
11,
12] and the quality of life questionnaire EQ-5D, consisting of five questions to measure health outcome. Preference-based utility values were applied [
13].
The overall influence on quality of life was analysed using EQ-5D utility values at three, six and 12 months. In a secondary analysis, a hand surgeon (HER) mapped DASH scores to EQ-5D dimensions at 12 months. DASH-derived EQ-5D utility values were compared to direct EQ-5D based values in an exploratory analysis.
Patients were classified as “returned to work” when they had resumed at least some labour market activity.
Costs
Three measures of costs were analysed; costs within the health-care sector, costs due to lost production and total costs (the sum of previous two costs). Costs within the health-care sector arise from diagnostic procedures, operations and rehabilitation and were calculated using the administrative prices paid by a referring hospital to our department (year 2009).
Costs due to lost production were calculated from the number of days of sick leave and the average earnings of the profession [Statistics Sweden;
http://www.scb.se] plus the payroll following the human capital approach [
14].
All costs are here presented in Euros and we used year 2009 average exchange rate for conversion (EUR 1 = SEK10.6213; [The Swedish Central Bank;
http://www.riksbank.se]).
The ethical committee of Lund University approved the study.
Statistical methods
Descriptive results are presented as median (minimum-maximum) values. The
χ
2-test was used to compare groups and the Kruskal-Wallis test were used to analyse costs by place of injury and by attachment to the labour market. The association between the final duration of sick leave and the DASH score at 3, 6 and 12 months, respectively, was analysed by the Kruskal-Wallis test. Pair wise correlation coefficients were estimated between DASH and EQ-5D at study measurements, and between their changes over time. Time to return to work was analysed by Cox-regression analysis stratifying for severity of injury (major injury/severe injury). Factors associated with health-care costs and total costs including costs of lost production were analysed using regression analysis [
15].
Discussion
In the present study, with randomly included patients with major or severe hand injuries (HISS score >100; and range 50–100, respectively) and while median HISS scores did not differ by place of injury, work-related injuries had highest total costs. Interestingly, the health care costs of the work-related injuries were 40% higher than those occurring at home or during leisure time controlling for age, DASH at injury and quality of life measured by EQ-5D at 12 months. The total costs were 34% higher in the present patients injured at work. Irrespective of place of injury, costs of lost production constituted the major part of total costs in the patients that worked full time before injury. In contrast, health care costs were +66% in patients working part-time before the injury. We can only speculate as to why in our sample, but it is possible that jobs, where people work part-time, are more likely to be manual and with a possibility of subsequent difficulty to return to work as well as involve greater risks of injury. In addition, the regression analysis also showed that the health care costs were higher for patients with an age > 50 years (+52%). Thus, we had independent cost increases from age and labour market attachment. At one year after injury, 17 patients were still on sick leave and the return to work was related to severity of injury. Patients with severe injuries were near as four times as likely to return to work at a given point in time compared to patients with major injuries (Figure
3). However, several factors may be involved in an injured patient´s ability to return to work and the interconnection of the factors is complex. In a study of costs and utility in the treatment of back and neck pain, the best predictor for return to work when comparing several commonly used health measures was in fact the EQ-5D [
16]. Another important factor for return to work is the patients’ sense of coherence in the recovery after hand injuries [
17,
18] and musculoskeletal pain [
19]. We continuously follow our patients with a tentative follow up of at least five years, since the importance of costs and other factors for hand injuries are high-lighted in literature [
6,
20‐
23].
The injuries of the present patients could essentially be divided in three equal types; amputations (29%), complex injuries (i.e. combination of fractures, tendon, vessel and nerve injuries; 40%) and major nerve injuries, including full house, in the forearm (29%), while a burn injury was only present in a single patient. The health care costs were higher in the patients with amputations and complex injuries than in the patients with nerve injuries, in spite of those full house injuries were included in the latter category. There was a need for microsurgical reconstruction of blood vessels in almost half of all the cases, which may increase health care costs. Interestingly, all cases with amputations needed microsurgery, while only 3/18 cases (17%) of the complex injuries required microsurgery. In accordance with previous reports [
5], the health care costs were only a minor part of the total costs for nerve injuries (i.e. > 80% were lost production). The corresponding percentages of lost production were even lower for the other two main categories (amputations 51%; complex injuries 69%). In the light of these results showing proportionately high costs of lost production, especially for nerve injuries, cost-efficient interventions leading to increasing return to work should have priority [
8]. The percentage of lost production (i.e. sick leave) is relevant when considering in on which patients resources for rehabilitation and return to work should be directed. The major part of total costs for nerve injuries were lost production indicating that the return to work should have priority in that category of patients. Nevertheless, the top priority of action is prevention of hand injuries, and this study supports interventions concentrated not only against work-related injuries, but particularly also towards injuries occurring during leisure, e.g. during “doing-your-self activities” [
7].
All patients were considered to be fully mobilized at three months and allowed to use their hands without restrictions and thus we report DASH at 3, 6 and 12 months. Patients with longer sick leave had higher DASH scores, but DASH scores improved (i.e. decreased) significantly over time in all patients (Table
2). The improvement of DASH from 3 to 12 months was clinically relevant (median score improvement >10 [
24]). In accordance with DASH scores, EQ-5D values also improved (i.e. increased) over time. The findings that patients who rated high EQ-5D values at 12 months had lower total costs are logical, since they had returned to work. There was a tendency of a positive association between increased improvement in DASH score between 3 and 12 months and total costs. One explanation may be that patients with high initial DASH score had a greater potential for improvement, but also were probably more costly in terms of health care resource need and sick leave.
A limitation of this study was that we randomly selected patients with a severe or major hand injury and not all patients during the study period. Our goal was to include as many patients as possible during the study period. These patients came randomly during day/night, weekdays and months and we (i.e. the authors) were also dependent on colleagues for including patients. Sometimes patients were not asked by the hand surgeon on call to participate in the study. However, we could not detect any differences in age, gender, weekday or month of injury between the included and not included patients (Table
1). However, there was a slight overrepresentation of major injuries among the included patients. The strength is the meticulous follow up of patients of costs of injury as well as the other aspects of outcome, such as the influence of sense of coherence [
17,
25].
We used an episode-perspective to assess the size of costs and to discuss associations between costs, place of injury and patient reported outcomes. It is not a traditional cost-of-illness study, which calculates the burden of illness born by health care and other sectors, but typically it lacks more detailed patient information, including outcomes of treatment. Thus, our approach provides more information of direct clinical interest.
Conclusions
We conclude that hand injuries at work were most costly both in terms of health care and costs of lost production. This finding is interesting since the severity, i.e. HISS, of work-related injuries did not differ from injuries occurring at home or during leisure and also in view of the fact that work environments have constantly improved over the last decades.
Consent
Written informed consent was obtained from the patient for publication of this report and any accompanying images.
Acknowledgements
The project was supported by grants from the Swedish Research Council (Medicine), Promobilia, Fanny Ekdahls stiftelse, HKH Kronprinsessan Lovisas förening för barnasjukvård, Region Skåne and Funds from the Skåne University Hospital Malmö, Sweden.
Competing interests
The authors declare that they have no competing interests related to the article.
Authors’ contributions
RC and LD designed the study, RC and HER collected all data, RC, HER, KSC and LD analysed data and all authors participated in the interpretation of data and writing the manuscript. All authors approved the final manuscript.