Cost-effectiveness analysis of routine follow-up for hip fracture patients after dynamic hip screw and intramedullary nail fixation

Highlights

• •Determining the appropriate outpatient management for patients following IM Nail or DHS for neck of femur fractures should be directed by clear objectives.
• •Routine post-operative surgical encounters with these patients rarely alter their clinical management.
• •We propose more selective follow-up protocols and adequately resourcing virtual alternatives.
• •The current outpatient hip fracture care pathway should be optimised to focus on comprehensive multidisciplinary treatment such as rehabilitation, fracture prevention and bone health optimisation.

Abstract

Introduction

The rising number of hip fractures has incentivised several quality improvement initiatives aimed at improving outcomes. These include the national hip fracture audit and the best practice tariff. Whilst there is an established standard of care for inpatients, the optimal outpatient management of patients after hip fracture fixation remains undefined. We aim to evaluate the cost-effectiveness of routine surgical appointments and provide evidence to support an improved outpatient hip fracture care pathway that focuses on more comprehensive multidisciplinary treatment.

Methods

This study retrospectively examined all patients who underwent either dynamic hip screw or intramedullary nail fixation for hip fragility fracture at our hospital over a 3-year period. Data was obtained from the Irish Hip Fracture Database (IHFD), the national integrated imaging system (NIMIS) and the medical charts. The number of outpatient appointments, postoperative radiographs and estimated cost of these patient encounters was examined. The rate of revision surgery as well as the type and timing of these interventions was also analysed.

Results

We included 272 patients. The mean number of inpatient and outpatient radiographs was 1.13 and 1.54 per patient respectively. There were 428 outpatient appointments scheduled, an average of 1.6 (SD 1.5) per patient. The median length of stay was 12 days (IQR 8-17). Only 16 (5.8%) patients had abnormal X-ray findings and only 8 (2.9%) patients underwent revision surgery. The majority (15/16, 94%) of patients with surgical complications presented with significant symptoms. There was a 13.1% DNA rate resulting in a cost to the health service of €1,400. The combined total cost of follow-up for all patients over the 3 year period was €92,252.

Discussion

Routine postoperative surgical encounters rarely alter a patient's management and contribute significantly to healthcare costs, resources and time loss for patients and their carers. We propose more selective follow-up protocols, adequately resourcing virtual alternatives and an improved outpatient hip fracture care pathway that focuses on comprehensive multidisciplinary treatment such as rehabilitation, fracture prevention and bone health optimisation.

Introduction

Hip fractures represent a substantial workload and economic cost to Irish hospitals. National literature reports an incidence of 407 per 100,000 in females and 140 per 100,000 in males aged over 50 years with neck of femur fractures accounting for 50% of fractures in those aged over 65 [1,2]. There are approximately 3700 neck of femur fractures in Ireland annually costing an estimated €45 million in acute hospital care. Of these patients, 43% are managed by Dynamic Hip Screw (DHS) or Intramedullary nail (IM Nail) fixation, 47% are managed by hemiarthroplasty and 4% undergo Total Hip Replacement (THR) [1]. Furthermore, the incidence of hip fractures is expected to increase by another 150% by 2046 as result of our ageing population [2]. As these numbers rise, so will the need for surgical intervention and outpatient services, causing considerable strain on a healthcare system with limited resources.

In Ireland, the introduction of the Irish Hip fracture Database (IHFD) and the Best Practice Tariff (BPT) has incentivised the optimisation of the inpatient management of patients suffering hip fractures. These initiatives have improved and standardised outcomes for this cohort. However, despite multiple advances in hip fracture care pathways, there is still no consensus on the appropriate outpatient management of hip fracture patients after internal fixation. Patients who have suffered low energy hip fractures are often fragile, co-morbid and in need of benevolent care. Outpatient appointments can be poorly tolerated and associated with patient/carer stress and the need for transport resources from long term care facilities. It is important to optimise the benefit of each outpatient visit.

The Irish National Clinical Programme for Trauma and Orthopaedic surgery has published an integrated care pathway (ICP) for hip fracture patients [3]. The current ICP guidance is to routinely follow-up patients 6 weeks after DHS or IM Nail fixation and 3 months after bipolar hemiarthroplasty (BPH) or THR in trauma. The AO foundation recommends follow-up of proximal femur fractures managed by internal fixation at 6 weeks and subsequently every 6 weeks until the fracture has healed [4]. However, there is generally a variety of individual practices among surgeons regarding routine radiographic and clinical follow-up.

This study was designed to evaluate the cost-effectiveness of routine orthopaedic outpatient and radiographic follow-up for hip fracture patients after DHS and IM nail fixation. Our aim is to optimise the outpatient hip fracture care pathway. We hypothesise that surgical consultations rarely change a patient's clinical course and that resources could be redistributed towards comprehensive multidisciplinary interventions such as primary prevention of fractures, postoperative rehabilitation, nutrition, falls prevention, bone mineral density assessment and pharmacological management of osteoporosis.

Methods

This study was undertaken at an Irish tertiary referral trauma unit with a catchment area of over 400,000 people. We included all patients who underwent DHS or IM Nail (short and long) for fractured neck of femur over a 3-year period between 1^st January 2017 and 31^st December 2019. Patients were excluded if they underwent prophylactic fixation for malignancy or were followed-up in a different institution. Data was extracted from the Irish Hip Fracture Database (IHFD) and radiographs were reviewed on the national integrated medical imaging system (NIMIS). Outpatient department (OPD) attendance times and dictated clinic letters were ascertained from the integrated patient management system (iPMS) and medical charts.

The IHFD is a clinically led, web-based national audit, which measures the care and outcomes of all people over the age of 60 with hip fractures. It is a collaborative venture, supported by the Irish Institute of Trauma and Orthopaedic Surgery (IITOS) and the Irish Gerontological Society (IGS), whilst the National Office of Clinical Audit (NOCA) provides operational governance. All 16 hospitals with orthopaedic services managing hip fracture patients submit to the database. Entries are made by dedicated audit coordinators through the IHFD Portal and linked with a hospital admission episode. The variables collected for each patient included age, gender, fracture type, method of internal fixation, ASA grade, anaesthetic type, pre-operative mobility, length of stay, admission date, type of facility the patient was admitted from, discharge date and where the patient was discharged to. Pre-operative mobility was measured using the “New Mobility Score” (NMS) (Table 1). This is a validated method of ascertaining a patient's functional ability indoors, outdoors and while shopping. It is out of a total score of 9, which equates to independence in all domains [5].

Table 1.

New Mobility Sore (NMS). A validated method of ascertaining a patient's functional ability indoors, outdoors and while shopping. It is out of a total score of 9, which equates to independence in all domains.

Mobility	Unable	Assistance of one person	With aid	Independent
Pre-fracture indoor walking	0	1	2	3
Pre-fracture outdoor walking	0	1	2	3
Pre-fracture shopping	0	1	2	3

The number outpatient visits per patient, the number of appointments that patients Did Not Attend (DNA) and the cumulative cost of these appointments was analysed. We also collected the number of inpatient radiographs, the number of outpatient radiographs, the timing of these radiographs and whether or not there were any abnormalities present. Finally, the 30-day revision rate, total revision rate and time to revision surgery was calculated. All outpatient visits took place in “face to face fracture clinics” with an estimated cost of €129 per appointment [6]. The estimated cost of a 2-view hip radiograph is €55 at our institution and the estimated cost of a DNA is €25 per appointment. It must be noted that the exact costs of these interactions are difficult to quantify as they encompass staffing costs and daily running costs. There are also indirect costs associated with these interventions such as the cost to society due to reduction in productivity.

An abnormal radiograph was defined as: lag screw cutting out, loss of fracture reduction, implant breakage, >10° varus collapse, >10mm of femoral neck shortening, delayed union at 12 weeks, periprosthetic fracture and auto-dynamisation during fracture healing. Of note, simple migration of the lag screw within the femoral head or a fracture that was noted to have united in “acceptable” varus, valgus, or rotation at routine followup was not classified as a fixation failure [7].

Continuous variables are displayed as median (interquartile range) or mean (standard deviation), whereas categorical variables are displayed as number and percentage. One-way ANOVA with Games Howell post hoc testing was used to compare the means of categorical variables. Data analysis and graphical presentation were performed using SPSS version 25.

Results

Our analysis included 272 patients with fractured neck of femur managed by either DHS or IM Nail. There were 103 patients in the DHS group, 114 patients in the Short IM Nail group and 55 patients in the Long IM Nail group. The median age was 83 years (IQR 74 - 88). Our cohort included 73 (26.8%) males and 199 (73.2%) females. There were 224 (82.4%) intertrochanteric, 36 (13.2%) subtrochanteric and 12 (4.4%) undisplaced intracapsular fractures. Seventy-nine percent of patients were either ASA II or ASA III grading. Patients typically had reduced mobility with a mean NMS of 6 (Interquartile range, IQR 3-9) and only 84 (30.8%) were suitable for discharge directly home after their acute hospital stay despite 223 (81.9%) living at home at the time of admission. The median length of stay (LOS) per patient was 12 days (IQR 8-17) (Table 2).

Table 2.

Patient Demographics LOS = Length of stay, LTC = Long term care

	DHS	Short IM Nail	Long IM Nail	Total
Patients (n)	103	114	55	272
Age (median, IQR)	82 years (71-87)	85 years (78-90)	78 years (69-87)	83 years (74-88)
Gender (Male:Female)	27:76	29:85	17:38	73:199
Fracture type (n, %)
- Intertrochanteric	93 (90.3%)	107 (93.9%)	24 (43.7%)	224 (82.4%)
- Subtrochanteric	0 (0%)	5 (4.4%)	31 (56.4%)	36 (13.2%)
- Intracapsular (undisplaced)	10 (9.7%)	2 (1.8%)	0 (0%)	12 (4.4%)
ASA grade (n, %)
- ASA 1	5 (4.9%)	4 (3.5%)	1 (1.8%)	10 (3.7%)
- ASA 2	39 (37.9%)	33 (28.9%)	17 (3.1%)	89 (32.7%)
- ASA 3	43 (41.7%)	55 (48.2%)	28 (50.1%)	126 (46.3%)
- ASA 4	5 (4.9%)	8 (7.0%)	1 (1.8%)	14 (5.1%)
- not documented	11 (10.7%)	14 (12.3%)	8 (14.5%)	33 (12.1%)
Anaesthetic type (n, %)
- General anaesthetic	16 (15.5%)	15 (13.2%)	14 (24.4%)	45 (16.5%)
- Spinal anaesthetic	85 (82.5%)	94 (82.4%)	38 (69.1%)	217 (79.8%)
- General + regional block	1 (1.0%)	4 (3.5%)	1 (1.8%)	6 (2.2%)
- General + spinal anaesthetic	1 (1.0%)	1 (0.9%)	2 (3.6%)	4 (1.5%)
Pre-op mobility (median, IQR)	6 (4-9)	5 (3-9)	7 (4-9)	6 (3-9)
LOS (median, IQR)	12 days (7-17)	12 days (8-18)	13 days (10-17)	12 days (8-17)
Admission from (n, %)
- Home	85 (82.5%)	90 (78.9%)	48 (87.2%)	223 (81.9%)
- Nursing Home / LTC	15 (14.6%)	17 (14.9%)	5 (9.1%)	37 (13.6%)
- Another acute hospital	3 (2.9%)	7 (6.1%)	2 (3.6%)	12 (4.4%)
Discharged to (n, %)
- Home	37 (35.9%)	25 (21.9%)	22 (40%)	84 (30.8%)
- Convalescence/Rehabilitation	38 (36.9%)	50 (43.9%)	24 (43.6%)	112 (41.2%)
- Return to LTC	15 (14.6%)	17 (14.9%)	5 (9.1%)	37 (13.6%)
- New admission to LTC	6 (5.8%)	11 (9.6%)	1 (1.8%)	18 (6.6%)
- Other acute hospital	3 (2.9%)	5 (4.4%)	1 (1.8%)	9 (3.3%)
- Hospital mortality	4 (3.9%)	6 (5.3%)	2 (3.6%)	12 (4.4%)

Current volume and cost of routine OPD and radiographic follow-up

Among the 103 patients in the DHS group, a total of 224 postoperative radiographs were performed and a total of 175 OPD appointments were scheduled. This corresponded to a mean of 2.17 radiographs and 1.7 OPD appointments per patient resulting in a total cost of €35,520.

The 114 patients in the short IM Nail group underwent a total of 247 postoperative radiographs and 126 OPD visits. The mean number of radiographs and OPD visits per patient were 2.16 and 1.13 respectively. The total healthcare cost of these resources for this group was €30,289. There were significantly fewer OPD appointments attended by patients in the short IM Nail group vs the DHS group. (p < 0.01)

The 55 patients in the long IM Nail group underwent a total of 177 postoperative radiographs and were scheduled for 127 OPD appointments. This resulted in a mean of 3.22 radiographs and 2.3 OPD appointments per patient at a total cost of €26,443. Their mean number of OPD appointments attended was significantly higher than the short IM Nail group (p < 0.001). There was no statistical difference intheir mean OPD attendance compared to the DHS group. (p = 0.1)

The total number of DNAs was 56/428 appointments giving DNA rates of 11% for the initial OPD scheduled visit and 9.6% for subsequent visits. DNAs in our study resulted in a cost to the health service of €1,400. The combined total cost of follow-up of all patients over the 3 year period was €92,252. (Fig. 1) (Table 3) (Fig. 2)

Fig. 1.

Number of inpatient and outpatient radiographs performed per patient journey (including those who had abnormal radiographs and those who required revision surgery). The median number of inpatient and outpatient radiographs performed per patient was 1.13 and 1.54 respectively.

Table 3.

a. Radiographs: Number and cost of inpatient and outpatient radiographs post-internal fixation for fractured neck of femur. b. Outpatient visits: Number and cost of outpatient visits post-internal fixation for fractured neck of femur.

	DHS	Short IM Nail	Long IM Nail	Total
Radiographs
Patients (n)	103	114	55	272
Inpatient radiographs per patient	1.24 (0.6)	1.0 (0.35)	1.23 (0.5)	1.13 (0.5)
(mean, SD)
Inpatient radiographs (total)	125	114	66	305
Outpatient radiographs per patient	1.51 (1.3)	1.19 (1.1)	2.03 (1.5)	1.54 (1.4)
(mean, SD)
Outpatient radiographs (total)	99	133	111	343
Abnormal radiographs	8	3	5	16
(Inpatient and outpatient)
Estimated total cost of post-op radiographs	€ 12,320	€ 13,585	€ 9,735	€ 35,640
Outpatient (OPD) visits
OPD visits per patient	1.7 (1.8)	1.13 (1.2)	2.3 (1.6)	1.6 (1.5)
(mean, SD)
OPD visits attended (total)	175	126	127	428
DNA first visit (n, %)	16 (15.5%)	8 (6.8%)	6 (10.9%)	30 (11%)
DNA subsequent visit (n, %)	9 (8.7%)	10 (8.8%)	7 (9.7%)	26 (16.6%)
Estimated total cost of post-op	€ 23,200	€ 16,704	€ 16,708	€ 56,612
OPD visits/DNA

DNA = Did Not Attend

Fig. 2.

Number of outpatient visits per patient and number of DNAs at each visit. DNA = Did Not Attend.

Revision surgery for complications

There were 16 abnormal radiographs in our study out of the total 648 performed. This included 8 patients in the DHS group, 3 in the short IM Nail group and 5 in the long IM Nail group. All the abnormal radiographs in the DHS group demonstrated either lag screw migration/cut-out (n=6) or varus collapse of the femoral neck (n=2). All of these patients had symptomatic complications and five of the six patients with lag screw cut-out required revision surgery. One patient had complete cut-out but was deemed too medically unfit and unsuitable for revision surgery.

Of the 114 patients who underwent short IM Nail fixation only 2 patients required revision surgery, one washout and debridement for early infection and another for revision of the distal locking screw. Two other patients were monitored closely as outpatients for lag screw migration, which remained stable.

Within the long IM nail cohort, 5 patients had an abnormal radiograph with only one patient underwent revision to BPH for lag screw cut-out at 16 weeks post operatively. Two patients developed malunion that was first identified at 3 months follow-up and one patient experienced loss of fracture reduction that subsequently healed in an adequate position. Our study identified no pattern in the timing to surgical complications and only two patients underwent revision within 30 days of surgery. There was no association between revision surgery and mortality (p = 0.34). (Table 4, Fig. 3)

Table 4.

Characteristics of abnormal XR and revision patients, BPH = Bipolar Hemiarthroplasty. THR = Total Hip Replacement. DB-THR = Diaphyseal-bearing Total Hip replacement.

	DHS	Short IM Nail	Long IM Nail
Patients with abnormal follow-up radiographs (n)	8	3	5
Time to first abnormal X-ray - Screw migration/Cutting-out	5 days 5 days 6 weeks 6 weeks 20 weeks 40 weeks	Cut-out monitoring From 2 days Cut-out monitoring From 6 weeks	8 days
- Varus collapse	4 weeks, 8 weeks
- Malunion / Non-union			3 months, 3 months
- Fracture displacement			9 days
- Autodynamisation		6 weeks	5 days
Re-operation (after 30 days)	1	1	0
Re-operation rate (total)	5	2	1
Clinically symptomatic with abnormal radiographs (n, %)	8 (100%)	3 (100%)	4 (80%)
Time to revision surgery	BPH Day 16 THR 5 weeks THR 20 weeks THR 22 weeks DB-THR 50 weeks	Washout and debidement 1 week Revision of distal locking screw 6 weeks	BPH 16 weeks

Fig. 3.

(A, B, C) Radiographs demonstrating a DHS complicated by lag screw cut-out and revised to a Total Hip Replacement (THR). Radiological complications typically presented within 8 weeks in our cohort.

Discussion

Determining the appropriate outpatient management for patients following IM Nail or DHS for neck of femur fractures should be directed by clear objectives. There should be a focus on optimising mobility, independence and quality of life after surgery. Our results suggest that routine postoperative surgical encounters and radiographs in this cohort rarely alter a patient's management and may represent misdirected health service expenditure that could be better invested in more comprehensive multidisciplinary care. In the context of fragility fractures, this means concentrating on the primary prevention of fractures, postoperative rehabilitation, nutrition, fall prevention, bone mineral density assessment and pharmacological management of osteoporosis.

Outpatient resource allocation and management has become an essential aspect of the provision of healthcare. Its importance was highlighted by the COVID-19 pandemic where an increasing number of patients were required to attend virtual consultations to reduce traditional clinic volumes. Furthermore, there are currently over 650,000 patients in Ireland awaiting an outpatient appointment, an increase of 51,000 (+8.5%) in the past 12 months [8]. Addressing this continuous rise in demand and improving patients access to tertiary care remains one of the priorities of our health service. Recently, increased consideration has been given to defining the most appropriate follow-up for various patient cohorts such as determining which patients are suitable for virtualised or “as required” outpatient care and which patients are most suitable for community-based follow-up. With the implementation of the Irish health's service new policy “Slaintecare”, our country will aim to restructure how and where patients are cared for with a focus on adequately resourcing primary and community care systems to support acute hospital services.

In Ireland, it is not routine practice to review patients who undergo hemiarthroplasty in outpatient fracture clinics. For patients who have undergone internal fixation, the current protocol in our hospital is to perform one post mobilisation radiograph and one outpatient radiograph at a 6-week interval with follow-up as required afterwards. However, routine outpatient follow-up of patients contributes significantly to healthcare costs, resources and time loss for patients and their carers. There is recent evidence to suggest selective follow-up of patients with hip fractures after internal fixation may be an appropriate alternative [9], [10], [11].

Our research found that patients attend an average of 1.5 outpatient appointments and receive 1.54 post-operative radiographs at a cost of €184 per visit (consultation with AP + Lateral radiograph). The total cost of follow-up was €92,252 over 3 years at our hospital alone. Despite this, only 8 patients (2.9%) underwent revision surgery. Furthermore, nearly all of the patients who required revision surgery had significant symptoms suggestive of a surgical complication. This demonstrates that routine surgical follow-up is not cost effective. Kuorikoski et al. report similar results with 0.23% of 428 scheduled outpatient appointments within 10 weeks of hip fracture fixation resulting in a change to the management plan. They also found patients with acute problems after hip fixation seek medical attention outside of the scheduled follow-up appointment [10]. Similarly, Halonen et al. noted only 0.9% of 995 patients after IM nail for trochanteric fracture required a change in management as a result of findings from routine follow-up [11]. The burden of routine outpatient appointments for patients and carers should also not be underestimated. The median age in our cohort was 83 years and 51.4% had severe systemic disease (ASA 3+). Sixty-five percent were not suitable for discharge home and may have required carer/family support or medical transport to attend the clinic appointment.

More selective surgical follow-up would also likely reduce the current DNA rate (13.1%), which was associated with substantial healthcare expenditure (€1,400). This was often due to death, inability to attend due to medical condition/transport limitations or a perceived lack of benefit after a prolonged inpatient stay. It was also common for patients who did not attend their first appointment to be routinely scheduled for another follow up appointment that they did not attend.

The results of our study also dispute the routine use of inpatient post mobilisation radiographs. Patients undergo surgery guided by intraoperative image intensifiers that save the final implant position and fracture reduction. Weight bearing status or implant related concerns should be addressed at this stage. Later radiographs to assess fracture union or subclinical loss of reduction are not neccesary as these are seldom indications for revision surgery in this frail cohort of patients [11], [12], [13]. We also suggest that patients who require non-routine follow-up for histology results or incremental advancements in weight bearing status can be followed by virtual alternatives whenever possible. Increasing virtualisation of care has been shown to reduce cost, consultation time, patient travel time and has been associated with high patient satisfaction. [14], [15].

The reoperation rate for IM Nail and DHS in this cohort was 2.9%. The IHFD report 2019 states the current reoperation rate for all neck of femur surgery including arthroplasty is 1%. Similar to other studies, we found it was common for patients to be symptomatic if they suffered complications that required revision surgery. As a result, when these patients presented to the hospital for assessment, it was often outside of their routine, scheduled visits. Our study found no patterns with regard to the timing of surgical complications [16], [17]

Strengths and limitations

This is the first study examine the cost effectiveness of routine follow-up for hip fracture patients in an Irish population. We used the well-regarded IFHD as a source of our data. This information proves timely during a period where the care pathways for hip fracture patients are being optimised. There is also an urgent need to reduce the outpatient waiting times in Irish hospitals.

The main limitation of this study is its retrospective nature. As a result outpatient notes and medical records informed our data. This risks the possibility of detection bias since patient data is not always properly documented. It must also be noted that the exact costs of these interactions are difficult to quantify as they encompass staffing costs and daily running costs. There are also indirect costs associated with these interventions such as the cost to society due to reduction in productivity.

Conclusion

We propose that patients are selectively followed up after undergoing hip fracture fixation with DHS or IM Nail. There may also be a role for virtualised appointments in certain settings. This modification to the outpatient care pathway may allow redistribution of resources towards other multidisciplinary, community-based care modalities.

Patients and their carers should avail of early supported discharge and should be assigned a “physio-buddy” to monitor their rehabilitation progress. Education should be provided on how to recognise the common complications of hip fracture surgery and a point of contact in the orthopaedic department should be offered. The potential for public health nurse wound supervision should be utilised in the community in conjunction with clear wound instructions.

Future research studies should also concentrate on defining the patient cohorts that require selective follow-up after internal fixation. Further studies should define the appropriate care pathways of patients following total hip arthroplasty in trauma as these patients present with different set of postoperative risks and outcomes.

Funding statement

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Ethical approval

Our study was conducted in accordance to the Helsinki Declaration as revised in 2013 and patient-identifying information was not presented in this report.

Declaration of Competing Interest

The authors have no conflicts of interest to declare.