Abstract
Background:
Coronavirus disease 2019 (COVID-19) has caused substantial disruptions to orthopaedic and trauma services. The purpose of the present study was to quantify its impact on surgical volume, hospitalizations, clinic appointments, and accident and emergency attendances to guide staffing and resource deployment for the sustenance of emergency services.
Methods:
Data were retrieved from all 43 Hong Kong public hospitals and 122 outpatient clinics from a population of 7.5 million residents. The “COVID-19 cohort” of patients who received treatment from January 25 to March 27, 2020, was compared with the “control cohort” of patients who received treatment during the same time of year over the past 4 years. Primary outcomes consisted of changes in patient diagnoses, number of operations performed, and hospitalizations during the COVID-19 pandemic. Secondary outcomes included differences in patient age and comorbidity, the nature of operations performed, types of anesthesia for orthopaedic procedures, difference in anesthetic times, wait times, and personal protective equipment (PPE) reserves.
Results:
A total of 928,278 patient-episodes (32,613 operations, 97,648 hospital admissions, 302,717 accident and emergency attendances, and 495,300 outpatient clinic attendances) were analyzed. Orthopaedic operations were reduced by 44.2%, from a mean (and standard deviation) of 795 ± 115.1 to 443.6 ± 25.8 per week (p < 0.001), with the ratio of emergency to elective operations increasing from 1.27:1 to 3.78:1. Operations for the treatment of upper and lower-limb fractures decreased by 23% (from 98.5 ± 14 to 75.9 ± 15.2 per week; p < 0.001) and 20% (from 210.6 ± 29.5 to 168.4 ± 16.9 per week; p < 0.001), respectively, whereas elective joint replacement and ligamentous reconstruction procedures decreased by 74% to 84% (p < 0.001). Operations for orthopaedic infections such as necrotizing fasciitis and septic arthritis remained similar (p > 0.05). The number of hospitalizations decreased by 41.2% (from 2,365 ± 243 to 1,391 ± 53 per week; p < 0.001), whereas clinical outpatient visits decreased by 29.4% (from 11,693 ± 2,240 to 8,261 ± 1,104 per week; p < 0.001). Patients did not endure longer wait times for emergency operations and accident and emergency consultations (p > 0.05). PPE consumption did not exceed procurement, with net increases in PPE reserves.
Conclusions:
Demand for orthopaedic care remains, despite weekly reductions of 351 orthopaedic operations, 974 hospital admissions, and 3,432 clinic attendances. Orthopaedic surgeons and health-care professionals should factor this into consideration during staffing and resource deployment.
Level of Evidence:
Prognostic Level III. See Instructions for Authors for a complete description of levels of evidence.
The response to the coronavirus disease 2019 (COVID-19) pandemic has necessitated strategic deployment of scarce health-care resources, especially in publicly funded health-care systems. Musculoskeletal conditions accounted for 602.3 million medical consultations, 2.2 billion prescriptions, and 21.5 million hospital discharges annually in the United States before the COVID-19 pandemic1. Orthopaedics straddles the worlds of both elective and emergency clinical services; as a result, the impact of COVID-19 on epidemiology and service demands, and the effectiveness of response measures, must be gauged for the purpose of guiding appropriations in staffing and equipment in order meet the challenges presented by COVID-19 while sustaining an emergency orthopaedic service.
We hypothesized that the COVID-19 pandemic has resulted in reductions in orthopaedic service provision in terms of surgical volume, hospitalizations, clinic appointments, and accident and emergency attendances, with more pronounced reductions in elective than emergency clinical services.
Materials and Methods
This epidemiological retrospective cohort study was carried out in compliance with the principles outlined in the Declaration of Helsinki. Ethics approval was granted by the Institutional Review Board of the University of Hong Kong/Hospital Authority Hong Kong West Cluster (institutional review board reference number: UW 20-275).
Our study was conducted in Hong Kong, which has a publicly funded health-care system serving 7.5 million residents2, with the public sector tasked to manage all suspected and confirmed COVID-19 patients. On top of managing all cases of COVID-19 in the region, public hospitals are tasked with managing all-cause emergency and elective inpatient and outpatient services, constituting almost 90% of hospital bed-days and virtually all critical emergencies in the region3.
Anonymized patient data were collected from all 43 Hong Kong public hospitals and 122 outpatient clinics in the region. Medical records, including inpatient hospitalizations, outpatient clinic attendances, surgical operations performed, and accident and emergency department attendances were retrieved from the territory-wide clinical data repository. Diagnoses and procedures were encoded according to the International Classification of Diseases, 9th Edition, Clinical Modification (ICD-9-CM). Comorbidity was represented according to the Charlson Comorbidity Index (CCI)4.
The study region reported its first confirmed COVID-19 case on January 23, 2020, with escalation to the “Emergency” response level on January 25, 2020, by the Hong Kong Government and Hospital Authority, the quasi-governmental organization managing all public hospitals in the region5. Government policies on social distancing included territory-wide school suspension, closure of theme parks, and cancellation of major social and sporting events6. Public hospital directives included reduction of elective orthopaedic operations by 80%; this directive mostly involved joint replacement, ligamentous reconstruction, benign skin lesion excision, and decompression of neuropathies (e.g., carpal tunnel release). Decisions on operative deferral were made in conjunction with treating physicians. These measures were taken with the aim of increasing admission capacity for COVID-19 cases, preserving personal protective equipment (PPE), and alleviating the pressure on intensive-care facilities. Emergency and semi-urgent operations such as fracture fixation were allowed to proceed, with regular review from hospital committees contingent on the latest epidemiological projections.
Universal clinical screening for COVID-19, which involved body temperature checks, history taking for COVID-19 symptoms, travel history, and contact history, was carried out for all orthopaedic patients. Physical examination with chest auscultation was performed and chest radiographs were made for patients requiring surgery and those with suspicious clinical findings. Severe acute respiratory syndrome (SARS)-CoV-2 reverse transcription-polymerase chain reaction (RT-PCR) testing of deep throat saliva or combined nasopharyngeal flocked swabs targeting the E gene was carried out in patients with suspicious clinical findings, with a low testing threshold.
In the private sector, some hospitals required negative SARS-CoV-2 RT-PCR test results before hospital admission and operation, whereas others were less stringent and only required absence of symptoms, travel history, and contact history before consultation and treatment. There was no legal restriction on the type and volume of surgical procedures performed in private hospitals.
The “COVID-19 cohort” consisted of patients receiving treatment in the 9-week period from January 25 to March 27, 2020, with follow-up data up until April 3, 2020. The “control cohort” consisted of patients with hospital admissions, orthopaedic operations, and outpatient clinic attendances from the same time of year in the past 4 years to account for annual variations (i.e., from January 25 to March 27 in 2016 and from January 25 to March 28 in 2017, 2018, and 2019 [for the same number of days to account for non-intercalary common years]). To mitigate variations of clinical service and demand due to different days of the week, a calendar week (7 days) was used as the unit of analysis.
Primary outcomes consisted of differences between the COVID-19 and control cohorts in terms of operations (elective and emergency), operatively treated orthopaedic emergencies (e.g., fractures, spinal cord injuries, necrotizing fasciitis), inpatient hospitalizations, and outpatient clinic attendances.
Secondary outcomes comprised differences in the number of operations performed in different age groups (pediatric and adolescent patients <18 years of age and geriatric patients ≥65 years of age), patient comorbidity as represented by the CCI, types of anesthesia for orthopaedic procedures, difference in anesthetic times, and wait times for emergency orthopaedic surgical procedures (defined as the time elapsed between emergency operation booking and the start of the operation). Data related to the composition of hospitalized orthopaedic inpatients, including the nature of admission (elective versus emergency), age, comorbidities, principal diagnoses, discharge rates and destinations, accident and emergency attendances, and wait times (defined as the time elapsed between registration and when the patient was first assessed by a registered medical practitioner), were also analyzed. Hospital admission criteria were not modified during the study period. For instance, patients with low back pain with “red flag” signs (e.g., fever, neurological deficit, sphincter disfunction, history of trauma, or history of malignancy) or suspicious radiographic findings (e.g., fracture or new osteoporotic vertebral collapse, pedicle destruction, end-plate erosion with disc destruction) were hospitalized for further investigation and management.
The levels of PPE reserves were reported as an indirect measure of attempts at conservation through reduction of elective clinical services.
To mitigate confounding effects of financial factors in health-seeking behavior, such as the pursuit of health-care services in the private sector, subgroup analysis was performed. Patients receiving public financial assistance could not afford private health care. Patients receiving public financial assistance were fully subsidized for treatment in public health care.
Statistical analysis was performed with use of SPSS for Windows (version 24; IBM). The Mann-Whitney U test was used for continuous parameters, and the chi-square test was used for categorical parameters. The level of significance was initially set at p < 0.05, with Bonferroni correction to adjust for multiple testing.
Results
A total of 928,278 patient-episodes, which included 32,613 operations (3,992 in the COVID-19 cohort and 28,621 in the control cohort), 97,648 hospital admissions (12,516 in the COVID-19 cohort and 85,132 in the control cohort), 302,717 accident and emergency attendances (36,744 in the COVID-19 cohort and 265,973 in the control cohort), and 495,300 outpatient clinic attendances (74,352 in the COVID-19 cohort and 420,948 in the control cohort), were included in analysis.
Orthopaedic Operations
The total number of orthopaedic operations performed during the COVID-19 pandemic decreased by 44.2% as compared with the control period (from 795 ± 115.1 to 443.6 ± 25.8 per week; p < 0.001). This finding was observed across all age groups, including those <18 years of age (56.9% reduction; p < 0.001), and to a lesser extent, those ≥65 years of age (35.4% reduction; p < 0.001) (Table I). Operatively managed patients in the COVID-19 group were slightly older (mean age and standard deviation [SD], 62.8 ± 22.5 compared with 59.6 ± 22.2 years; p < 0.001) and had more medical comorbidities as reflected by the CCI (p < 0.001) than those in the control group.
TABLE I.
Differences in Orthopaedic and Trauma Operations During COVID-19 Pandemic
| COVID-19 Cohort (2020) Weekly Total | Control Cohort (2016-2019) Weekly Total | Change | P Value | |||
| Mean ± SD | Median (IQR) | Mean ± SD | Median (IQR) | |||
| Operations | ||||||
| Total (n = 32,613) | 443.6 ± 25.8 | 446 (437-455) | 795 ± 115.1 | 829.5 (795-859) | −44.2% | <0.001 |
| Elective (n = 13,440) | 92.8 ± 30.6 | 84 (74-109) | 350.1 ± 80.8 | 380.5 (344.5-393) | −73.5% | <0.001 |
| Emergency (n = 19,173) | 350.8 ± 27.1 | 342 (337-363) | 444.9 ± 47.8 | 447.5 (421-470.5) | −21.2% | <0.001 |
| Pediatric and adolescent patients (<18 years old) (n = 1,796) | 19.4 ± 4.3 | 18 (17-22) | 45 ± 9.4 | 46.5 (39-49) | −56.9% | <0.001 |
| Geriatric patients (≥65 years old) (n = 14,792) | 228.4 ± 23.5 | 223 (221-239) | 353.8 ± 45.9 | 366.5 (342.5-381.5) | −35.4% | <0.001 |
| Traumatic fractures and dislocations | ||||||
| Upper-limb fractures (n = 4,229) | 75.9 ± 15.2 | 76 (65-82) | 98.5 ± 14 | 101.5 (89.5-108.5) | −23% | <0.001 |
| Lower-limb fractures (n = 9,098) | 168.4 ± 16.9 | 172 (164-179) | 210.6 ± 29.5 | 205 (190-221.5) | −20% | <0.001 |
| Hip fractures (n = 6,192) | 113.3 ± 17.2 | 113 (109-120) | 143.7 ± 21.9 | 136.5 (128-156.5) | −21.2% | <0.001 |
| Pelvic fractures (n = 306) | 7.2 ± 3.6 | 8 (4-10) | 6.7 ± 3.3 | 6 (4-9) | 7.5% | 0.665 |
| Joint dislocations (n = 677) | 11.4 ± 3.3 | 11 (9-14) | 15.9 ± 4.7 | 16 (13-19.5) | −28.3% | 0.009 |
| Spinal fractures (n = 231) | 3.3 ± 2.1 | 2 (2-4) | 5.6 ± 2.7 | 5 (4-7.5) | −41.1% | 0.02 |
| Spinal pathologies | ||||||
| Spinal cord compression (n = 35) | 2.2 ± 1.4 | 2 (1-3) | 1.3 ± 0.5 | 1 (1-1.5) | 69.2% | 0.111 |
| Spinal cord injury (n = 52) | 1 ± 0 | 1 (1-1) | 1.8 ± 1.1 | 1 (1-3) | −44.4% | 0.105 |
| Central cord syndrome (n = 35) | 1 ± 0 | 1 (1-1) | 1.5 ± 0.8 | 1 (1-2) | −33.3% | 0.464 |
| Infections | ||||||
| Necrotizing fasciitis (n = 219) | 3.6 ± 1.8 | 3 (2-5) | 5.3 ± 3.4 | 5 (2-7.5) | −32.1% | 0.247 |
| Septic arthritis (n = 456) | 9.1 ± 3.6 | 10 (6-13) | 10.4 ± 3.4 | 10 (8-13) | −12.5% | 0.442 |
| Osteomyelitis (n = 409) | 7.6 ± 3.1 | 7 (6-9) | 9.5 ± 3.4 | 9.5 (7-11) | −20.0 % | 0.146 |
| Bone and soft-tissue malignancies (n = 221) | 3.2 ± 2.1 | 3 (2-4) | 5.3 ± 2.5 | 5 (3.5-7) | −39.6% | 0.03 |
| Elective reconstructive operations | ||||||
| Total hip arthroplasty (n = 382) | 2.6 ± 2.4 | 2 (1-3) | 10 ± 3 | 10 (8-12.5) | −74.0% | <0.001 |
| Total knee arthroplasty (n = 2,132) | 11.4 ± 12.9 | 4 (2-28) | 57 ± 16.4 | 62.5 (50.5-65) | −80.0% | <0.001 |
| Anterior cruciate ligament reconstruction (n = 553) | 2.4 ± 2.2 | 1 (1-3) | 14.9 ± 4.6 | 15.5 (12.5-18) | −83.9% | <0.001 |
Reduction in surgical volume was not proportionate across operative categories: whereas elective operations decreased by 73.5% (from 350.1 ± 80.8 to 92.8 ± 30.6 per week; p < 0.001), emergency operations decreased by only 21.2% (from 444.9 ± 47.8 to 350.8 ± 27.1 per week) (Fig. 1, Table I). The majority of orthopaedic operations performed during the COVID-19 pandemic consisted of emergency operations, with an emergency:elective ratio of 3.78:1 (compared with a pre-pandemic ratio of 1.27:1).
Fig. 1.
Significant decreases (p < 0.001) were observed in both elective procedures (73.5% decrease) and emergency procedures (21.2% decrease) during the COVID-19 pandemic. Box = IQR, horizontal line = median, x = mean, whiskers = variability outside the upper and lower quartiles, and points outside lines or whiskers = outliers.
The numbers of operatively treated upper and lower-limb fractures decreased by 23% (p < 0.001) and 20% (p < 0.001), respectively. The number of operatively treated hip fractures decreased by 21.2% (p < 0.001). Whereas reductions also were seen in the numbers of operatively treated spine fractures and joint dislocations, the differences were nonsignificant after Bonferroni correction for multiple testing (p > 0.001). The number of operatively treated pelvic fractures, spinal emergencies (including spinal cord injuries, spinal cord compressions, and central cord syndromes) remained similar (Table I).
Emergencies related to orthopaedic infections remained unabated: the numbers of operations for the treatment of necrotizing fasciitis, septic arthritis, and osteomyelitis were not significantly different between the cohorts. Operations for the treatment of malignant bone and soft-tissue lesions decreased from 5.3 ± 2.5 to 3.2 ± 2.1 per week (p = 0.03), which was not significant after Bonferroni correction (Table I).
Conversely, the numbers of joint replacement operations plummeted: total knee replacements decreased by 80.0% (p < 0.001), whereas total hip replacements decreased by 74.0% (p < 0.001). Anterior cruciate ligament reconstructions decreased by 83.9% (p < 0.001) (Table I). The weekly breakdown of the numbers of operations is shown in Table II. A lag time of 2 to 3 weeks, from the time of initiating emergency response measures to actual reductions in elective operations, was observed.
TABLE II.
Orthopaedic Surgical Volume in Weeks Preceding and During COVID-19 Pandemic
| Period and Week | Upper-Limb Fracture Fixation | Lower-Limb Fracture Fixation | Total Hip Replacement | Total Knee Replacement | Anterior Cruciate Ligament Reconstruction | Elective Operations | Emergency Operations | Total Operations |
| Before beginning of local COVID-19 pandemic without emergency response measures | ||||||||
| 23 December-29 December 2019 | 79 | 153 | 9 | 83 | 15 | 405 | 433 | 838 |
| 30 November-6 December 2019 | 109 | 149 | 15 | 70 | 9 | 399 | 473 | 872 |
| 7 December-13 December 2019 | 74 | 187 | 18 | 71 | 17 | 384 | 494 | 878 |
| 14 December-20 December 2019 | 73 | 177 | 13 | 79 | 13 | 391 | 478 | 869 |
| 21 December-27 December 2019 | 78 | 114 | 9 | 33 | 7 | 187 | 402 | 589 |
| 28 December 2019-3 January 2020 | 67 | 171 | 12 | 50 | 6 | 228 | 448 | 676 |
| 4 January-10 January 2020 | 91 | 170 | 15 | 79 | 16 | 391 | 446 | 837 |
| 11 January-17 January 2020 | 80 | 145 | 16 | 74 | 14 | 416 | 408 | 824 |
| 18 January-24 January 2020 | 108 | 149 | 15 | 63 | 18 | 336 | 490 | 826 |
| After beginning of local COVID-19 pandemic with emergency response measures | ||||||||
| 25 January-31 January 2020 | 47 | 146 | 8 | 28 | 7 | 124 | 331 | 455 |
| 1 February-7 February 2020 | 65 | 115 | 4 | 12 | 1 | 109 | 338 | 447 |
| 8 February-14 February 2020 | 69 | 128 | 2 | 31 | 3 | 153 | 337 | 490 |
| 15 February-21 February 2020 | 65 | 119 | 1 | 4 | 1 | 75 | 342 | 417 |
| 22 February-28 February 2020 | 90 | 127 | 2 | 0 | 1 | 59 | 386 | 445 |
| 29 February-6 March 2020 | 57 | 140 | 2 | 1 | 0 | 74 | 363 | 437 |
| 7 March-13 March 2020 | 60 | 141 | 1 | 0 | 0 | 66 | 391 | 457 |
| 14 March-20 March 2020 | 63 | 123 | 1 | 2 | 1 | 84 | 362 | 446 |
| 21 March-27 March 2020 | 60 | 101 | 0 | 2 | 3 | 91 | 307 | 398 |
Patients undergoing emergency operations did not experience increased wait times (134.7 ± 135.3 compared with 136.4 ± 213.1 minutes; p = 0.323). The time for general anesthesia inclusive of intubation increased by 18.0% (from 16.7 ± 12 to 19.7 ± 13.8 minutes; p < 0.001). Despite attempts to switch from general anesthesia to regional anesthetic techniques, which do not generate aerosols, the proportion of patients receiving general anesthesia increased slightly (from 75.4% to 79.6%; p < 0.001).
Accident and Emergency Attendance for Orthopaedic Conditions
The number of emergency visits fell by 44.7% from 7,388 ± 643 to 4,083 ± 587 per week (p < 0.001). Reductions were more prominent in the pediatric and adolescent age group (69.4%; p < 0.001) than the geriatric group (40.3%; p < 0.001) (Table III).
TABLE III.
Differences in Orthopaedic Trauma Accident and Emergency Attendances and Outpatient Clinic Attendances During COVID-19 Pandemic
| COVID-19 Cohort (2020) Weekly Total | Control Cohort (2016-2019) Weekly Total | Change | P Value | |||
| Mean ± SD | Median (IQR) | Mean ± SD | Median (IQR) | |||
| Accident and emergency orthopaedic attendances | ||||||
| Total (n = 302,717) | 4,082.7 ± 586.7 | 4,208 (3,684-4,648) | 7,388.1 ± 643.1 | 7,547 (6,969-7,868.5) | −44.7% | <0.001 |
| Pediatric and adolescent attendances (<18 years old) (n = 21,754) | 171.9 ± 31.5 | 172 (161-179) | 561.3 ± 103.7 | 581.5 (479.5-639) | −69.4% | <0.001 |
| Geriatric attendances (≥65 years old) (n = 79,926) | 1,153 ± 115.9 | 1,172 (1,085-1,248) | 1,931.9 ± 144.8 | 1,915.5 (1,863.5-2,033) | −40.3% | <0.001 |
| Traffic-related trauma (n = 13,161) | 227.2 ± 28.9 | 220 (201-250) | 308.8 ± 33.3 | 310 (289.5-333) | −26.4% | <0.001 |
| Industrial-related trauma (n = 30,938) | 411.6 ± 101 | 447 (333-481) | 756.5 ± 144 | 805.5 (713-852.5) | −45.6% | <0.001 |
| Domestic-related trauma (n = 46,188) | 686.4 ± 80.9 | 672 (636-746) | 1,111.4 ± 106.4 | 1,098 (1,066-1,131.5) | −38.2% | <0.001 |
| Sports-related trauma (n = 12,666) | 130 ± 52.7 | 137 (83-171) | 319.3 ± 67.8 | 329.5 (273.5-369) | −59.3% | <0.001 |
| Hospital admissions (n = 67,573) | 1,084 ± 101.2 | 1,093 (1,050-1,162) | 1,606 ± 122.5 | 1,624 (1,564-1,684.5) | −32.5% | <0.001 |
| Outpatient clinic attendances | ||||||
| Total (n = 495,300) | 8,261.3 ± 1104.4 | 8,357 (8,015-8,979) | 11,693 ± 2,240.3 | 12,285.5 (11,925.5-12,796.5) | −29.3% | <0.001 |
| New attendances (n = 60,639) | 1,153.1 ± 509.6 | 1,030 (992-1,134) | 1,396.1 ± 272.2 | 1,475 (1,442-1,536) | −17.4% | 0.005 |
| Follow-up attendances (n = 434,661) | 7,108.2 ± 874.5 | 7,157 (6,984-7,838) | 10,296.9 ± 1,974.9 | 10,788 (10,470.5-11,294.5) | −31.0% | <0.001 |
The number of sports-related trauma cases decreased significantly by 59.3% (p < 0.001). Significant reductions (p < 0.001) also were seen in trauma related to traffic (26.4%), industrial (45.6%), and domestic causes (38.2%) (Fig. 2). Wait times were reduced from 91 ± 102 to 54 ± 66 minutes (p < 0.001).
Fig. 2.
Significant decreases (p < 0.001) were observed in accident and emergency-related attendances for trauma and injuries involving sports (59.3%), industrial (45.6%), domestic (38.2%), and traffic-related causes (26.4%) during the COVID-19 pandemic. Box = IQR, horizontal line = median, x = mean, whiskers = variability outside the upper and lower quartiles, and points outside lines or whiskers = outliers.
Outpatient Clinic Attendance
The number of outpatient visits decreased by 29.3% (from 11,693 ± 2,240.3 to 8,261.3 ± 1,104.4 per week; p < 0.001) (Fig. 3). Follow-up visits, constituting the majority of outpatient clinic appointments, decreased by 31.0% (p < 0.001), whereas new attendances decreased by 17.4% (p = 0.005) (Table III).
Fig. 3.
A significant decrease (p < 0.001) was observed in outpatient clinic attendances (29.3%) during the COVID-19 pandemic. Box = IQR, horizontal line = median, x = mean, whiskers = variability outside the upper and lower quartiles, and points outside lines or whiskers = outliers.
Hospital Admissions
Inpatient hospitalizations decreased by 41.2% (from 2,364.8 ± 243.1 to 1,390.8 ± 53.3 per week; p < 0.001). Whereas elective admissions decreased by 58.9% (from 918.1 ± 167.1 to 377.6 ± 59.4 per week; p < 0.001), emergency admissions decreased by only 30.0% (from 1,446.6 ± 119.4 to 1,013.1 ± 90.4 per week; p < 0.001) (Fig. 4). Pediatric admissions decreased by 59.1% (p < 0.001), whereas geriatric admissions decreased by 36.7% (p < 0.001) (Table IV). Patients admitted during the COVID-19 pandemic had more comorbidities (median CCI, 0 [interquartile range (IQR), 0 to 2] compared with 0 [IQR, 0 to 0] in the controls; p < 0.001). The top 3 principal diagnoses for admissions during the COVID-19 pandemic were hip fractures (104.1 ± 18 weekly admissions), low back pain (74.44 ± 12.12 weekly admissions), and distal radial fractures (66.8 ± 11.2 weekly admissions). Subgroup analysis of patients receiving social security financial assistance revealed no significant change in the number of emergency hospitalizations (p = 1.0).
Fig. 4.
Significant decreases (p < 0.001) were observed in elective (58.9%) and emergency (30%) hospital admissions during the COVID-19 pandemic. Box = IQR, horizontal line = median, x = mean, whiskers = variability outside the upper and lower quartiles, and points outside lines or whiskers = outliers.
TABLE IV.
Difference in Orthopaedic Hospitalizations During COVID-19 Pandemic
| COVID-19 Cohort (2020) Weekly Total | Control Cohort (2016-2019) Weekly Total | Change | P Value | |||
| Mean ± SD | Median (IQR) | Mean ± SD | Median (IQR) | |||
| Total (n = 97,648) | 1,390.8 ± 53.3 | 1,392 (1,337-1,435) | 2,364.8 ± 243.1 | 2,438.5 (2,316-2,503) | −41.2% | <0.001 |
| Elective (n = 36,451) | 377.6 ± 59.4 | 346 (340-437) | 918.1 ± 167.1 | 975 (933.5-1,009.5) | −58.9% | <0.001 |
| Emergency (n = 61,197) | 1,013.1 ± 90.4 | 1,036 (983-1,094) | 1,446.6 ± 119.4 | 1,466 (1,393.5-1,529) | −30% | <0.001 |
| Pediatric and adolescent admissions (<18 years old) (n = 6,114) | 63.1 ± 8.8 | 64 (59-68) | 154.1 ± 28.5 | 160.5 (133.5-170) | −59.1% | <0.001 |
| Geriatric admissions (≥65 years old) (n = 47,397) | 719.6 ± 29.6 | 712 (701-731) | 1,136.7 ± 118.1 | 1,143.5 (1,096-1,224.5) | −36.7% | <0.001 |
| Patients receiving public financial assistance (n = 22,208) | 443.7 ± 18.3 | 448 (434-452) | 506 ± 176.5 | 478.5 (335.5-682) | −12.3% | 1 |
Personal Protective Equipment (PPE) Reserves
The reduction in elective clinical services, in concert with PPE procurement, ensured increasing reserves despite increasing numbers of suspected and confirmed COVID-19 cases7. According to weekly communications from the Hospital Authority head office, the cumulative amount of PPE reserves at >100 institutions for >67,000 staff included 25 million surgical masks (compared with 18 million 2 months previously), 2.9 million isolation gowns (compared with 2.2 million 2 months previously), 3.7 million face shields (compared with 0.5 million 2 months previously), and 2.1 million N95 respirators (compared with 1.1 million 2 months previously)8 (Fig. 5).
Fig. 5.
Despite PPE utilization for emergency services, conservation by reducing elective services and acquisition of PPE from manufacturers avoided depletion of PPE reserves. (Source of data: Hospital Authority COVID-19 Bulletin Issues 17-37.)
Discussion
Many professional organizations have published empirical guidelines on adjusting clinical services during the COVID-19 pandemic9-12. It is imperative that decisions made by administrators and hospital leadership be backed by clinical data in order to provide realistic projections and to better inform judicious deployment of staffing and resources. Decisions based on clinical data could avoid overestimating the effectiveness of emergency measures and jeopardizing the provision of emergency services.
The numbers of hospital admissions and surgical procedures for the treatment of fractures of both the upper and lower limbs decreased, with reductions seen across all causes of trauma including sports, traffic, industrial, and domestic. Legally enforced social distancing measures, including school suspension, limiting public gatherings, and work from home and “stay home” initiatives6, could have been contributing factors. Restriction of mobility could have impacted the risk of fracture as the majority of falls occur outdoors13. Significant reductions were observed in the numbers of some but not all injuries, with dislocations, pelvic fractures, and spinal fractures among the exceptions. The incidence of severe orthopaedic infections remained stable.
Administrative measures played a direct role in service provision. Priority was given to the sustenance of emergency clinical services, including admissions and operations, with the provision of “elective-urgent procedures”14,15 inevitably affected. Reductions in elective service allowed conservation of PPE, reduction of unnecessary hospital traffic, and preservation of staffing to afford deployment to COVID-19 wards. Reduction in elective hospital admissions also vacated hospital beds and equipment (including ventilators) for the care of COVID-19 patients. Stable patients who were scheduled for clinic appointments were contacted via telephone by nurses and were advised not to return for follow-up unless absolutely necessary. Special arrangements for medication refill were in place as required.
Our findings confirmed that such measures effectively decreased surgical volume for elective procedures such as joint and ligamentous reconstruction by 70% to 90%. Reduction of elective surgical procedures also contributed to the decrease in the number of elective hospital admissions by 58.9%, which allowed conversion of orthopaedic wards into surveillance beds for suspected cases of COVID-19. Modest reductions (29.3%) were observed in outpatient clinic visits; despite prior telephone contact by nursing staff, many patients opted to return for their appointments, contributing to a lower-than-expected (31.0%) decrease for follow-up cases.
The implementation of emergency-response measures was not without difficulty: time and training were required for universal temperature checks at clinics, clarifications and recording of travel history, and collection of specimens for SARS-CoV-2 testing. The time taken for general anesthesia increased, likely for donning PPE during aerosol-generating procedures. Despite attempts to administer regional anesthesia in preference to general anesthesia, regional anesthesia was not possible for all patients because of comorbidities (e.g., therapeutic anticoagulants), anesthetic expertise, and patient preference. Expedient transferal to convalescent hospitals not initially designed for “semi-acute” patients necessitated additional medical and nursing staff during weekends and holidays. Substantial staffing resources were required to call back patients to reschedule clinic appointments. Infrastructural modifications such as conversion of orthopaedic wards into negative-pressure facilities required time and cost outlays before being able to accommodate COVID-19 patients. Orthopaedic staff volunteering for deployment to COVID-19 wards required training before undertaking duties beyond their clinical subspecialty.
Alterations in health-care-seeking behavior also could have contributed to changes in observed disease incidence. Patients may have delayed or opted not to seek medical attention for a number of reasons: fear of contracting COVID-19, altruistic consideration for health-care workers, financial constraints, or preference for private health-care facilities. To mitigate the confounding effects of the last 2 factors, we performed subgroup analysis on patients receiving public assistance. The number of emergency hospital admissions was not significantly different; thus, this factor needs to be factored in for health-care planning as it affects bed occupancy and PPE usage.
Our study, despite covering >900,000 patient episodes, was limited by data capture from public hospitals, which accounted for almost 90% of hospital bed-days in the region3; therefore, a minority of patients were not included in our analysis. Disease reporting was affected by changes in health-care-seeking behavior, despite attempts to adjust for patients seeking health care in the private sector.
Nevertheless, to our knowledge, this is the first quantitative study of the impact of COVID-19 on orthopaedic services. Our territory-wide multicenter database enabled data capture across 100 institutions serving a population of 7.5 million residents. We reported observations on changes in disease incidence during this COVID-19 pandemic and described the effect of central measures to restrict elective and nonemergency hospital services and their actual outcomes. This information could be critical for health-care planning, such that more accurate estimation of inpatient bed numbers, PPE conservation, and the number of staff available for redeployment can be modeled.
COVID-19 presents challenges to the medical profession and impacts orthopaedic and traumatology epidemiology and services. Despite attempts at reduction, there remains a substantial demand on orthopaedic and trauma care because of the special nature of our services. We hope that our findings will be useful to orthopaedic surgeons and health-care professionals in mounting a better-informed response toward the COVID-19 pandemic.
Supplementary Material
Footnotes
Investigation performed at Queen Mary Hospital, University of Hong Kong, Hong Kong SAR
Disclosure: This study was financially supported by the Sanming Project of Medicine in Shenzhen “Team of Excellence in Spinal Deformities and Spinal Degeneration Diseases” (SZSM201612055). The study sponsors had no role in the study design, data collection, analysis, interpretation, writing of the report, and submission decision. On the Disclosure of Potential Conflicts of Interest forms, which are provided with the online version of the article, one or more of the authors checked “yes” to indicate that the author had a relevant financial relationship in the biomedical arena outside the submitted work (http://links.lww.com/JBJS/F929).
References
- 1.United States Bone and Joint Initiative. The burden of musculoskeletal diseases in the United States (BMUS). 4th ed 2016. Accessed 2020 May 4 https://www.boneandjointburden.org/fourth-edition [Google Scholar]
- 2.Hong Kong Census and Statistics Department. Population - overview. 2020. April 9 Accessed 2020 Apr 12 https://www.censtatd.gov.hk/hkstat/sub/so20.jsp
- 3.Hong Kong Food and Health Bureau. Report of the strategic review on healthcare manpower planning and professional development. 2017. Accessed 2020 May 4 https://www.fhb.gov.hk/en/press_and_publications/otherinfo/180500_sr/srreport.html [Google Scholar]
- 4.Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40(5):373-1-9. [DOI] [PubMed] [Google Scholar]
- 5.Hong Kong Special Administrative Region. CE announces activation of Emergency Response Level in relation to novel coronavirus infection. 2020. January 26 Accessed 2020 Apr 12 https://www.info.gov.hk/gia/general/202001/26/P2020012600087.htm [Google Scholar]
- 6.Leung GM, Cowling BJ, Wu JT. From a sprint to a marathon in Hong Kong. N Engl J Med. 2020. April 30;382(18):e45 Epub 2020 Apr 15. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Centre for Health Protection. Latest situation of novel coronavirus infection in Hong Kong. 2020. Accessed 2020 Apr 12 https://chp-dashboard.geodata.gov.hk/covid-19/en.html [Google Scholar]
- 8.Hospital Authority Head Office Human Resources Division. COVID-19 bulletin. 2020. April 10 Accessed 2020 Apr 12 https://www.ha.org.hk/haho/ho/pad/200410c.pdf [Google Scholar]
- 9.Royal College of Surgeons. Guidance for surgeons working during the COVID-19 pandemic. 2020. March 20 Accessed 2020 Apr 12 https://www.rcseng.ac.uk/coronavirus/joint-guidance-for-surgeons-v1/ [Google Scholar]
- 10.American College of Surgeons. Clinical issues and guidance. 2020. Accessed 2020 Apr 12 https://www.facs.org/about-acs/covid-19/information-for-surgeons [Google Scholar]
- 11.British Orthopaedic Association. Management of patients with urgent orthopaedic conditions and trauma during the coronavirus pandemic. 2020. Accessed 2020 Apr 12 https://www.boa.ac.uk/uploads/assets/ee39d8a8-9457-4533-9774e973c835246d/COVID-19-BOASTs-Combined-v1FINAL.pdf [Google Scholar]
- 12.National Health Service. Clinical guide for the management of trauma and orthopaedic patients during the coronavirus pandemic. 2020. Accessed 2020 Apr 12 https://www.england.nhs.uk/coronavirus/wp-content/uploads/sites/52/2020/03/specialty-guide-orthopaedic-trauma-and-coronavirus-v1-16-march-2020.pdf [Google Scholar]
- 13.Li W, Keegan TH, Sternfeld B, Sidney S, Quesenberry CP, Jr, Kelsey JL. Outdoor falls among middle-aged and older adults: a neglected public health problem. Am J Public Health. 2006. July;96(7):1192-1-9 Epub 2006 May 30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.DePhillipo NN, Larson CM, O’Neill OR, LaPrade RF. Guidelines for ambulatory surgery centers for the care of surgically necessary/time-sensitive orthopaedic cases during the COVID-19 pandemic. J Bone Joint Surg Am. 2020. April 13 Epub 2020 Apr 13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.American College of Surgeons. COVID-19 guidelines for triage of orthopaedic patients. 2020. Accessed 2020 Apr 12 https://www.facs.org/covid-19/clinical-guidance/elective-case/orthopaedics [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.