Introduction: The removal of implants after fracture healing has always been a topical and controversial issue. Undisputed however, is the removal of implants which are associated with complications. In this paper, we evaluated the indications, the outcome and economic cost to the patient and the health system of routine removal of implants after union.
Patients and methods: This is a two-year comparative, prospective study carried out in three hospitals in North Central Nigeria. All patients who presented for implant removal and consented to the study were recruited. Fracture union was confirmed both clinically and radiological. Direct costs of treatment paid for by the patients and number of absences from work and school were used as economic indices.
Results: Forty-seven patients comprising 20(42.6%) males and 27(57,4%) females, mean age of 31.6 (±13.4) years were analysed. Forty-two (89.4%) had fixation with plates and screws, three (6.4%) had screws with cerclage wire and two(4.2%) had tension band wiring. Patient request was the main indication for removal in 34(72.3%), surgeon’s request without symptoms in 7(14.9%),six(10.7%) were symptomatic comprising four(8.5%) due to postoperative chronic osteomyelitis and intractable pain in two(4.3%). Total cost of implant removal was $33,293.59 ($ 708.37 ±22.10; range $366.97-1,100.92). Total cost of removal in patients with symptomatic implants-in-situ was $3,678,90( $ 613.15±14.50) One patient had a refracture. Mean duration away from work or school was 16 days.
Conclusion: Routine removal of implants after fracture union constituted great waste of highly needed cash in both direct cost and loss of man-hour especially in a dwindling economy in an era of evidence based Medicare and best practice. It is recommended that each hospital adopts strict criteria for implant removal.
The removal of orthopaedic implants after fracture has healed has always been a topical issue, firstly because the science of biomechanics of internal fixation is highly dynamic with development of newer and better fixation devices1 and secondly because the criteria for removal has never been clearly documented.2,3,4 What is however not contestable is the removal of implants in-situ which are beset with complications after the fracture has healed. These include pain around the region bearing the implants, post operative infection, broken implants and those adjacent vital structures.5,6,7 The removal are also not without their own problems and includes neurovascular injuries especially in the hands of junior members of the team, re-fracture, wound sepsis.6,7,8,9. Current literature does not appear to support routine removal to protect against allergy, carcinosis or metal detection.2,3,4,6,10
In children, it is generally believed that implants should be removed as soon as their functions have been achieved to avoid interfering with their growth. In most cases, this has not been the case unless those closely related to the growth plate. There is however no evidence in the current literature to support or refute the practice of routine implant removal in children.10,11,12,13 Podeszwa11 et al have suggested the use of bioabsorbable implants in children with physeal and epiphyseal fractures in a comparative study.
The aim of this study is to evaluate the indications, the outcome and the economic costs to the patient and the health system, the practice of routine removal of these implants.
This is a two year comparative and prospective study carried out in three different hospitals namely Jos University Teaching Hospital (JUTH) Jos, Bauchi State Specialist Hospital (BSSH), Bauchi, a secondary care institution and Daisyland Orthopaedic/Trauma hospital (DOTH) Jos, a privately owned hospital. All the authors work in these hospitals.
Patients data were entered into a proforma as they presented, these included age, sex, mechanism of initial injury, bone/bones fractured, date of fixation, the cost of initial operation, complications if any in the post fixation period, indications for removal, documented cost of removal, the duration of hospitalisation, period of absence from work and the patients satisfaction with the procedure. The cost of treatment was the actual amount paid for the hospital services and included admission cost, drugs and surgery fees.
Fracture union was confirmed both clinically and radiological. The patients had all been counselled earlier during the follow up after surgery for their initial fracture fixation that the implants could be left indefinitely if they so desired or removed after twelve and eighteen months for upper and lower limb fractures respectively on confirmation of fracture union.
The cost of lost man hour from work is difficult to assess especially as most of the patients are self-employed and children. We therefore decided to equate the number of days of absence from work as a function of the minimum wage in the country and each day lost to illness is considered as one point loss.
Indirect cost like transport of relatives looking after the patients, feeding away from home and miscellaneous costs cannot be estimated and considered as a constant for all the patients.
To evaluate the financial waste, the total money spent on the cost of removal and the total cost incurred by symptomatic patients were calculated separately. The difference forms the loss to the patients since this category of patients was operated even though considered not essential. The total man-hour is similarly calculated and analysed.
Over the period under review, 47 patients were recruited, nine (19.1%) from BSSH; 13(27.6%) from DOTH and 25 (57.6%) from JUTH. This represents 14.1% of the total of 332 major and intermediate orthopaedic operations carried out during this period in the three hospitals. All the patients that presented for implant removal and consented to the study were included. There were 20(42.6%) male and 27 (57.4%) female with a mean age of 31.6(± 13.4) years. Road traffic accident was the commonest cause of initial injury in 35(74.5%) cases with fall from heights and assaults in six (12.8%) and five (10.6%) respectively. The most common bone was femur in 20(42.6%), the distribution is as shown in table 1. Forty-two (89.4%) cases had internal fixation with plate/screws, three (6.4%) had screws and augmentation with cerclage wires and two (4.3%) with cerclage / K wires.
Patients’ request was the commonest indication in 34(72.3%) cases. The authors advised the removal in seven (14.9%) patients. Six (12.8%) patients had removal for specific indications and comprised post-operative wound infection in four (8.5%) patients and intractable peri-implant pain in two (4.3%) patients. All the infective cases were tibia fractures which were open fractures ab initio. The peri-implant pains occurred in one patella fracture that had tension band wiring and proximal femoral shaft fracture that had open reduction with plate and screws. This is depicted in Table II.
The mean cost of implant removal was $ 708.37 (±22.10; range $366.97-1,100.92) and the total cost for all the patients was $ 33,293.59. The average cost of removal in the symptomatic group of patients was $ 613.15 (±14.50) and the total cost was $ 3,678.90. The amount of money that would have been saved is $ 29,614.69 which is the cost of the money spent by the asymptomatic patients for unnecessary operations. This is shown in table III.
The mean duration of hospital stay for all the patients(n=47) was 2.6 (±1.01) days with the mode at 2days. The average time spent away from work is 15.4(±2.2; range of 7-21 days).The mean duration away from work was 16.3(±1.96; range 15-20) days and the mean duration of hospitalisation was 3.3(±1.6 days) for the patients with symptomatic implants in-situ .
The mean post-removal follow-up period was 13 (range11-24) months. One (2.1%) patient re-fractured through the previous fracture line. Forty-six (97.9%) patients had no complications.
In general, patients harbouring implants that have symptoms that could be traceable to the implants in-situ should always have them removed. Plates are stress shielding devices and are generally advised to be removed in the lower extremities. Intramedullary(IM) nails are stress sharing devices and will not cause osteopenia around the implants, therefore intramedullary nails can be left in situ without definite need for their removal, however if mal-alignment is encountered even in the presence of an IM nail, it should be removed to facilitate reconstruction at the later stage. Though complications are documented, when done correctly for appropriate indications, the benefits far outweigh the complications.. In an era of evidence based best practice, it is recommended that each hospital adopt criteria for the removal of implants as this would save for the patients huge financial resources which can be spent on themselves and their families and with the National Health Insurance badly needed resources for improved service delivery in these hospitals.
The indications for orthopaedic implants removal will always remain a subject of debate. Studies of outcome data are as varied.. In this study, patients with peri-implant pain, experienced relief from intractable pain following their operations. No intra-operative feature could explain their symptoms. Objectively, the visual analogue pain score improved from 6.3 preoperative to 1.0 at one month post operation and been pain free since. This finding was corroborated by Minkowitz et al who found a significant improvement in pain scores decreasing from mean of 5.5 to I.3 in their study. They also found an overall improvement of 76% at one year of follow up. This optimistic result was not reproduced as indicated in another report by Busam6 et al who warned that implant removal for pain was at best unpredictable and depends more on the type and location of the implant rather than the removal itself. Similarly, Morshed14 et al found in his study on paediatric femoral fractures that up to half of his patients had a non-debilitating pain at 2-5 years post-injury regardless of presence or absence of the implant. They were not sure what the cause of this could be. We believe that the relief in pain experienced has a lot to do with the psychology of the patients rather than actual pain due to the removed implant.
Removal of implant was therapeutic in treating chronic osteomyelitis with implants in-situ. A thorough debridement was undertaken and wounds irrigated with dilute hydrogen peroxide and saline. Discharges stopped on antibiotic therapy based on the culture and sensitivity pattern and dressings. This was not unexpected as implants act as foreign bodies which sustain the infection by the formation of a thin biofilm which antibiotics cannot penetrate to reach the organisms
The main indications for implant removal is the choice of patient most of which are females (23 females compared to 11 males). From our observations during the course of treatment, female patients allowed more psychological overlay in the perception of their implants-in-situ. This is further supported by the fact that all the 23 female patients expressed gains in the post operative period as being so much happier since the removal. This compared with only nine of the male patients. Two had not experienced any gains at all. It could also be deduced that men are more reluctant in choosing to electively remove their plate after healing because economic considerations had a more overpowering influence since these implants were no bother in situ anyway. On the other hand, the male sex was found to be an important significant factor in the likelihood of having an implant removed in a Canadian study by Sidky and Buckley..They believe this was due to other factors such as litigation. It is our firm belief that patient education with tact and patience will go a long way in preventing unnecessary operations for implant removal.
We advised removal in seven patients out of which five were children ten years and under. This was because we believed that implants in children should be removed as soon as the fractures healed to prevent implants being buried deep in the healing callus which make removal more difficult and hazardous much later age. This view is also shared by Kanlic and Cruz in their current concepts review of femur fractures in children and explained further that such implants could cause peri-prosthetic fractures and or eventually impede adult reconstruction in their old age.
The only complication in this series was a re-fracture in a seven year old which occurred exactly two weeks after removal of the implant. The patient had fallen in a jump landing awkwardly on the healed fracture limb. This has been well documented in literature with the incidence varying widely from 2.5% -40%. Studies using single photon absorptiometry showed cortical atrophy and decreased bone mineral density may be responsible for the refracture in these patients. The implants act as stress shields preventing normal bone metabolism. It was shown that this effect is negligible after six months. In retrospect, the plate and screws should have been left in situ for up to 12 months and some form of protection should have been advised in the immediate post-operative period.
The estimation of cost of treatment has always been difficult because some indirect costs can never really be assessed. Two aspects of cost estimation were assessed and include the actual cost incurred and paid for in the hospital and the number of days lost from work or school.
The total amount spent on removal was $ 33,293.59 by all patients and total spent by those with pre-removal symptoms was $3,678.90. This means that $ 29,614.69 was total amount that could have been saved if strict policy for implant removal in place. This was an enormous amount of money could have been spent on improving socioeconomic standards of the patients who had to pay from their own pockets.
Similarly, the total duration away from work is 723 days for all patients while symptomatic was 98 days. This means that 625 days could have been saved. This comprised 606 days of work loss by the workers and 92 days of school missed by the patients who are 18 years and below. These are difficult to convert to losses in terms of cash because the workers are in different employments and there is no uniformity in the salaries of workers in both the private and public sectors in Nigeria, nor could the amount of school work lost be assigned any monetary value.
Patients assessment of the operative procedure was taken into consideration and 98.8% of them were satisfied with the outcome even though would not want to undergo any further operative treatment. Two patients who were indifferent to the whole process were children and their fathers answered for them. Both complained of cost of removal while the father of the child with re-fracture wished we had delayed the removal.