This was a comparative-retrospective study, involving five secondary referral and one tertiary referral hospitals in Yogyakarta Special Province: Yogyakarta Regional Public Hospital, Wates Regional Public Hospital, Wonosari Regional Public Hospital, Panembahan Senopati Bantul Regional Public Hospital, Universitas Gadjah Mada Academic Hospital, and Dr. Sardjito General Hospital. This study was approved by the Medical and Health Research Ethics Committee Faculty of Medicine, Public Health and Nursing Universitas Gadjah Mada.
We included all patients who presented to the emergency departments or orthopedic clinics who were then diagnosed with new onset (< 1 week) fractures based on history clinical, and radiological examinations. The fractured bones included in this study were: clavicle, scapula, humerus, radius, ulna, hand, spine, pelvic, femur, tibia, fibula, and foot. For the COVID group (during the pandemic period), we included all patients who presented to the hospital from 20th March - 29th May 2020. For the control group, the period was from 1st April - 10th June 2019.
We excluded the patients with pathological fractures due to bone tumors (primary and metastatic), genetic bone abnormalities, unhealed previous fractures (poor union or non-union), periprosthetic fractures, and patients with data missing from their medical records.
The collected data included the sociodemographic characteristics of the patients (age and gender), mechanisms of injury, injury location, fractured bone, type of fracture (closed/open fracture), concurrent fracture, osteoporotic fracture, multiple trauma, hospital stay duration, referral, treatment, and patients who refused recommended surgery or inpatient care. The data were obtained from the medical records.
We categorized the mechanisms of injury as fall from standing height, fall from height <1 m, fall from height >1 m, bicycle injury (not involving collision with motor vehicle), motor vehicle accident/MVA (any injuries involving collision with motor vehicle, as the vehicle rider or hit by motor vehicle), others (any specific mechanisms of injury such as being hit, trapped, cut, machine-related, etc., not classified in the other groups), and unknown (unspecific mechanisms of injury). We considered fall from standing height, fall from height <1m, and the bicycle injury as low energy trauma; where fall from height >1m, MVA, and others as high energy trauma. The injury locations were grouped as at home (in the house/living place and the surroundings), road (including the sidewalks), others (school, workplace, sports fields, and other specified places), and unknown.
The fractured bone was documented as clavicle, scapula, humerus, radius and/or ulna, hand, spine, pelvic, femur, tibia and/or fibula, foot, and patella. The patients were stated to have concurrent fracture when they had more than one fracture based on the mentioned bone group. We considered a fracture as an osteoporotic fracture when it fulfilled all the following criteria: 1. The patient age was more than 50 years old (female)/more than 60 years old (male); 2. Occurred in proximal humerus, distal radius, thoracic/lumbar vertebrae, or proximal femur; and 3. Caused by a low energy trauma [17,18]. The patients were classified as having multiple trauma when they had any injuries involving more than one body region and required treatment from more than one specialist for the injuries [19].
We grouped the treatment as conservative and surgical treatment. For analyzing the treatment, we excluded patients who were referred to another hospital. The patient was classified to receive the surgical treatment for the specified fracture when they received the treatment involving incision with an instrument, performed by orthopedic surgeon, in the operating theatre.
The statistical analysis was performed using SPSS 23.0 (IBM Corp., Armonk, New York). We tested all the numeric variables with Kolmogorov-Smirnov test for the normality of distribution. The difference of mean was analyzed with t-tests for data with normal distribution, and with Mann-Whitney tests for data with abnormal distribution. For categorical variable, we conducted chi-squared and used Fisher exact tests when the expected count was less than five. The p-value of <0.05 was considered significant.