Abstract
Introduction
The Global Burden of Disease Study conducted in 2019 estimated an incidence of 14.2 million hip fractures with an associated years lived with disability of 2.9 million. 1 Yong et al reported an increase in numbers of hip fractures in Singapore between 2000 to 2017, 2 with Singapore ranked second highest in age- and sex-standardized incidence rate per 100,000 population. 3 The costs involved in the management of hip fractures are not limited to costs of acute hospitalisation, but also consist that of rehabilitation and discharge disposition. 4
Value in health care was originally defined as outcomes relative to cost, and value is only created if lower costs are accompanied by unchanged or improved outcomes. 5 Value-Based Healthcare was subsequently introduced as an extension to include concepts such as protocolised care, and is understood as the trade-off between outcomes and costs as a potential measure of the effectiveness of healthcare services. 6 To combat the exponentially rising costs in musculoskeletal surgeries in aging populations, protocolised care have been implemented in elective surgeries such as knee and hip arthroplasty.7,8 Protocolised care have been shown to improve financial outcomes whilst encouraging partnerships between care providers, ultimately improving patient outcomes as well as key performance indicators.9,10
Existing studies illustrate conflicting results after implementation of protocolised care for hip fracture patients: while some authors have shown that protocolised care results in reduced length of stay and lower readmission rates, with a greater proportion of patients discharged home, these have not been replicated in other studies.11,12 Positive results from protocolised care have not been consistently achieved in all institutions, with low-volume hospitals reporting higher complication rates and poorer outcomes.9,10,13 Institutions performing 50 or less cases per year were considered low volume while those performing 100 or more cases per year were considered high volume. 14 Moreover, the significant variability in terms of length of stay and comorbidity burden in hip fracture patients, as compared to total joint arthroplasty patients, makes cost containment difficult. 15
A multi-disciplinary protocolised care for hip fracture surgery was first implemented in our institution in 2017. This study aims to analyse its impact on key performance indicators including admission within 4 hours from the emergency department and time to surgery from admission within 48 hours (parameters adapted from the British Orthopaedic Association: Care of Patients with Fragility Fracture guidelines) 16 ; and acute outcomes of surgically managed hip fracture patients including post-operative length of stay, discharge disposition and 30-day inpatient mortality which have also been used in previous studies.17,18 We hypothesise that a multi-disciplinary protocolised care results in improved care indices and short-term outcomes following hip fracture surgery.
Methods
This study was reviewed by our institutional review board with waiver of consent granted as de-identified data was used in this study. A retrospective study on outcomes of patients with surgically treated low energy hip fractures was performed. We compared outcomes of patients managed prior and after the implementation of a multi-disciplinary protocolised care.
Inclusion criteria comprised patients aged 60 years and above, admitted for an acute isolated fracture of the hip (defined as the neck of femur, intertrochanteric or subtrochanteric region), sustained from a fall from standing height, and underwent surgical intervention for the index fracture during the acute admission. Patients excluded were those who sustained hip fractures from high energy trauma, polytrauma, patients with more than one fracture sustained at admission, pathological fractures, and atypical femur fractures. Atypical femoral fractures were excluded as they have a distinctively different pathophysiology, postulated to be associated with prolonged osteoclastic suppression.19,20
We extracted data from our institution’s electronic medical records: data extracted from January 2015 to January 2017 before the implementation of the multi-disciplinary protocolised care were grouped as the pre-protocol cohort (PRE-PROTOCOL) and data extracted from February 2017 to January 2020 after implementation of the multi-disciplinary protocolised care were grouped as the post-protocol cohort (POST-PROTOCOL).
Prior to the implementation of our institution’s protocolised care in 2017, there was a lack of multidisciplinary consolidated pathway from the Emergency Department to the acute hospital, the community hospital and subsequent secondary prevention. There was also no clearly defined strategies of pain management and protocol for venous thromboembolism prophylaxis. The involvement of geriatricians in the care of a hip fracture patient was on an ad hoc basis, with no coordinated input from geriatricians and anaesthetists to facilitate early access to surgery. Further, early mobilisation on the first post-operative day with a trained physiotherapist was not always guaranteed. Last, there was a lack of a robust partnership with a community hospital for step down care with an established safe discharge criteria including stable vital signs in an afebrile patient, no ongoing wound concerns, adequate analgesic control, no signs of venous thromboembolism and patient being able to ambulate with aids.
Surgeon seniority and training remained similar across both the PRE-PROTOCOL and POST-PROTOCOL periods. Implants already approved in the institutional tender were used throughout the PRE-PROTOCOL and POST-PROTOCOL periods and did not differ significantly. The majority of intra-capsular fractures underwent bipolar hemiarthroplasties whilst a mix of extra- and intra- medullary devices were used for peri-trochanteric hip fractures as appropriate.
The multi-disciplinary protocolised care envisaged the seamless transition of surgically treated hip fracture patients from presentation at the emergency department, coordinated multidisciplinary care in the acute setting, to continued rehabilitation in community hospitals, and return to the community with secondary prevention.
Since the implementation of the protocolised care in February 2017, adapted from the American College of Surgeons National Surgical Quality Improvement Program, 21 patients diagnosed with a hip fracture at the Emergency department (ED) would be assessed clinically to exclude any concomitant ongoing medical emergencies such as stroke or myocardial infarction that necessitate urgent intervention. Investigations for peri-operative assessment which include full blood count, electrolytes, an electrocardiogram, and a plain chest radiograph were performed at the ED, with optimisation commenced immediately.
A stepwise pain relief protocol as agreed upon by the anaesthetic pain team and geriatricians, was instituted, specifically avoiding non-steroidal anti-inflammatory drugs (NSAIDs). 500-1000 mg of paracetamol every 4-6 hourly was provided, with the use of oral opiates and intravenous morphine as additional analgesia. Femoral blocks were given by a trained anaesthetist if there was inadequate pain control at rest, and these were given either in the emergency department or the inpatient ward. Chemoprophylaxis for deep vein thrombosis with subcutaneous enoxaparin was started in the absence of contraindications such as head injuries or suspected gastrointestinal bleeds.
Early access to surgery was facilitated by prioritizing inpatient bed availability for hip fracture patients, a checklist (as fully endorsed by the Orthopaedic, geriatric, and anaesthetic teams) to fast-track surgically fit patients, access to early ancillary scans to confirm occult fractures of the hip not seen on initial radiographs, and a dedicated operative team for these patients. Within the first 24 hours of admission, all hip fracture patients were reviewed by a dedicated ortho-geriatric team, who co-managed these patients and minimised unnecessary pre-operative referrals to other specialties. Only cardiac conditions necessitating optimization such as new onset unstable angina, arrhythmias, murmurs, a recent established myocardial infarction, and heart failure were referred for cardiac assessment and optimisation. Based on established risk factors such as ischaemic heart disease, history of heart failure, stroke or transient ischaemic attacks, diabetes mellitus requiring insulin and renal dysfunction with a creatinine >170 umol/L, a cardiac risk stratification guideline was established for the orthopaedic team to counsel patient and family members with no need for specialist referrals.
A large majority of surgeries were performed by a senior Orthopaedic resident supervised by a fully trained Orthopaedic surgeon in dedicated trauma lists. Following early surgery, based on NICE guidelines to facilitate nursing and mobilisation, all patients were reviewed by trained physiotherapists for mobilisation on the first post operative day, including weekends and public holidays.
To minimise re-admission from the community hospitals, a safe transfer criteria including (i) absence of infection and fever, (ii) healthy post-surgical wounds with no erythema or drainage, (iii) adequate pain relief in order to effect transfers, and (iv) absence of urinary retention or unresolved constipation were established between the acute hospital and community hospitals. Visits by an advanced practice nurse to the partner community care hospitals ensured transfer of specialised nursing skills and served as a liaison with surgeons for any patient care issues to minimize unnecessary readmissions.
Potential confounding variables such as patient demographics, ASA scores, fracture type and type of surgical intervention, were obtained from the electronic medical records.
Acute hospital key performance indicators were compared between the pre- and post-protocol cohorts: admission within 4 hours from the emergency department, time to surgery from admission within 48 hours, acute hospital length of stay post-operatively within 9 days, discharge disposition and 30-day inpatient mortality. We also examined the rate of readmissions to the acute hospital from a community hospital.
Statistical analysis
Kolmogorov-Smirnov test of normality was used to assess the normality of the distribution of data. We used Pearson’s chi-squared test to compare categorical data, and student’s
Results
Our study recruited a total of 1636 patients: 701 patients were allocated to the PRE-PROTOCOL group and 935 patients to the POST-PROTOCOL group. Our data set was normally distributed (Age, df = 1636,
Demographic data in study cohort.
Clinical indices and types of surgeries in study cohort.
Risk reduction analysis of key performance indicators after implementation of protocolised care.
Mean time to surgery was shorter in the POST-PROTOCOL group (86.2 ± 94.1 vs 63.3 ± 59.9 hours; mean difference, 22.9; 95% CI, 15.4 to 30.4;
Mean post-operative length of stay was shorter in the POST-PROTOCOL group (15 vs 11 days; mean difference, 4; 95% CI, 3 to 5;
30-day inpatient mortality rates were not statistically dissimilar between the groups (0.4% vs 1.0%;
Discussion
The main finding in our study was that the implementation of a multi-disciplinary protocolised care for hip fracture surgery improved key performance indicators such as time to admission, time to surgery, as well as length of post-operative stay in the acute hospital, despite a cohort of patients with greater comorbidity burden. Though we found that patients in the POST-PROTOCOL group were statistically significantly older, the difference in mean age of 1.6 years is unlikely to be clinically significant.
Nayar et al showed that from 2011 to 2017, time to hip fracture surgery decreased by 2% annually, accompanied by decreased complication rates, 22 and a study by Uri et al has shown that with incentives, practical measures can be implemented for institutions to achieve shorter waiting time for hip fracture surgery. 23 Implementation of a multi-disciplinary protocolised care in our institution achieved a significant improvement in proportion of patients (36.8% vs 56.4%) being operated within 48 hours of admission, with mean time to surgery reduced by 22.9 hours. Common reasons for a delay in time to surgery can be divided into two domains: peri-operative optimisation such as correction of electrolyte imbalance or cessation of oral anti-coagulants, or pre-operative assessment including echocardiogram or reviews by other specialists. Though Sigemeth et al found no difference in mortality rates between patients who underwent surgery within or after 48 hours of admission, after adjusting for ASA, mental score, and mobility score, 24 most studies found that operative delay beyond 48 hours was associated with increased mortality.25–28 Other benefits of early hip fracture surgery include resultant shorter length of stay and resultant cost savings.24,29–31 However, it is important that in patients who are sicker and more frail, adequate time is spent in optimising their medical conditions pre-operatively. In our retrospective review, we discovered the trend that patients surgically treated for hip fractures in the POST-PROTOCOL group were of higher ASA status. It is now increasingly recognised that pain and immobility in non-operated hip fractures poses significant morbidity and mortality risks, and this has to be balanced against underlying co-morbidities determining fitness for surgery.32,33 Morrissey et al determined that the association between delay in surgery and mortality only became statistically significant beyond the 24-h mark, 34 indicating the balance to be achieved between peri-operative optimisation and early hip fracture surgery. We postulate that with a common consensus between Orthopaedic surgeons, anaesthetists and geriatricians, implementing a multi-disciplinary protocolised care provides greater accessibility to multidisciplinary care and better optimisation pre-operatively, hence even sicker patients could be operated on successfully. This is substantiated by comparable mortality rates when compared to the PRE-PROTOCOL group with better physiological status.
Shorter time spent in the ED has been traditionally considered as good hip fracture practice, with a study by Thiam et al showing that patients who spent more than 4 hours in the ED were more at risk of cardiac events and prolonged inpatient stay. 35 In our study, a greater proportion of patients were admitted to the ward within 4 hours (86.0% vs 97.2%) in the POST-PROTOCOL group. A study by Hughes et al showed that delays in admission result in longer stays in the acute hospital, leading to increased risks of malnutrition and nosocomial infections, as well as delayed patient movements within the hospital, leading to poorer health economics. 36 This finding was supported by Clague et al, who found that a delay in admission increased length of stay and postulated that patients deteriorate whilst waiting for admission. However, they also found that shorter stays in the ED may contribute to in-hospital mortality, though no causal relationship could be established. 37 A similar parameter studied by Vidal et al found that each day of delay from fracture to admission is associated with increased risk of in-hospital mortality as well as 1-year mortality, however this was mainly attributed by the access to healthcare or recognition of the severity of a hip fracture by patient or their caregivers. 38 In short, hip fracture patients should not experience any delay in admission in absence of a justifiable reason. This protocol led to a greater proportion of patients admitted within 4 hours, possibly by working with hospital administration to ensure a more streamlined admission process for these hip fracture patients.
Studies have shown that prolonged length of stay was associated with medical complications such as pressure ulcers, delirium, institutionalisation, and urinary tract infections, as well as reduced odds of successful rehabilitation. 36 In this study, reduction in post-operative length of stay in the POST-PROTOCOL group could be attributed to the streamlined workflow with the community hospitals to right site the patients to receive intensive rehabilitation in a timely manner at the step-down care facility. If otherwise, these patients would have stayed longer in the acute hospital. Although these patients’ acute medical and surgical issues have resolved, as evident in the lower rates of readmission from the community hospital, they were not able to manage at home and were not functionally safe for discharge. As the main focus for this group is rehabilitation, transfer to a community hospital with intense physiotherapy and occupational therapy is ideal, illustrating the application of right siting of hip fracture patients. Further, all patients enrolled in the protocol receive physiotherapist assessment the day after surgery unless medically contraindicated. With timely referral to allied health professionals, patients were mobilized early and that facilitated transfer to a step-down facility sooner. This is in agreement with Hughes et al, stating that the availability of step-down facilities and good support network can help to reduce the cost of prolonged acute hospital length of stay, which can contribute up to 84% of the total cost of an inpatient hip fracture cycle. 36 It is important to strive for early rehabilitation and efficient use of the acute hospital stay, but it is also imperative that we ensure this does not result in worse post-operative outcomes such as severe adverse events and death after hip fracture surgery, a possibility illustrated by Pascal et al. 39 A multi-disciplinary protocolised care allows patients to be mobilised early and right sited to partner step-down care facilities. The latter can safely and effectively continue post-surgical care by a mutual understanding of transfer criteria and continuum of rehabilitation protocols as determined in our protocolised care. This in turn also frees up inpatient beds for other patients requiring acute care, leading to more efficient utilisation of limited healthcare resources.
Though, a multidisciplinary protocol which includes more clinicians as well as allied health professionals theoretically incurs higher healthcare costs, the management of this subgroup of complex geriatric patients with multiple medical co-morbidities often involves input from other specialties at some point in the same acute hospital setting. Without protocolised care, such referrals are likely to be reactive in nature and will be delayed until a new medical complication is surfaced, resulting in either delayed time to surgery or increased length of stay in the hospital. With protocolised care, these complex patients are reviewed early, identifying potential non-orthopaedic issues and early preventative measures can be implemented. Though the aim of our study is not to specifically evaluate the financial benefits brought about by our implemented protocol, we expect the reduced time to surgery and length of hospital stay, as well as avoidance of downstream healthcare costs from prevention of medical complications to bring about potential financial savings.
Most patients in our geriatric population do not have private insurance and fall under this multi-disciplinary protocolised care as part of subsidised healthcare, adding to national healthcare costs rather than patient’s out-of-pocket hospitalisation fees. Hence, there is an impetus for public healthcare institutions to move patients downstream to reduce national healthcare costs by right-siting patients. Institutions, till date, have not been penalised for not meeting target outcomes, but these indicators are assessed every quarter and if necessary, quality improvement projects will be instituted.
In terms of sustainability of such a protocol, though there is no direct incentivisation given to the managing team, there are indirect benefits such as greater availability of beds for elective patients. This protocol is endorsed by our institution’s management, with buy-in from all stakeholders. For instance, the community hospital is our active stakeholder and works with the managing team to sustain seamless transfer of patients to stepdown care.
Limitations of this study include its retrospective nature, but it is difficult to perform a prospective randomised control trial in a vulnerable cohort of hip fracture patients. Further, the application of our results may be limited in other institutions or healthcare systems as this study was performed in a single institution. Finally, our study did not analyse the length of stay in step-down care facilities, cost data, functional outcomes, readmission and re-operations as these were outside the scope of our study. To our knowledge, this is one of the largest studies comparing the effects on key performance indicators after the implementation of a multi-disciplinary protocolised care for hip fracture surgery.
In conclusion, this study showcases the potential clinical incentives derived from a multi-disciplinary protocolised care for hip fracture surgery. A multidisciplinary approach to geriatric hip fractures in a high volume institution improves admission process and access to surgery, with resultant shorter hospital stays, and no increase in inpatient mortality.
