The study protocol and consent forms were approved by the Institutional Review Board (IRB), and registered in the Thai Clinical Trials Registry (registration no. TCTR20180302007). The study design and reporting format were based on Consolidated Standards of Reporting Trials (CONSORT) guidelines. Patients aged older than 50 years who were diagnosed with pertrochanteric hip fracture (either intertrochanteric or femoral neck fracture) during October 2016 to November 2017 were screened for the study. The exclusion criteria were patients with pathological fracture, defined as fracture in patients who had a history of any bone tumor either primary or metastasis. This condition was suspected in patients with osteolytic/osteoblastic lesion around the hip from preoperative radiographs. In addition, specimens from the fracture site was sent for pathological examination in most cases to confirm osteoporotic fracture, hypercalcemia (corrected total serum calcium more than 10.5 mg/dL), renal impairment with estimated glomerular filtration rate (eGFR) less than 30 mL/min/1.73 m2, abnormal liver function test with aspartate aminotransferase (AST) or alanine aminotransferase (ALT) level greater than twice the upper limit of the normal range, and previous vitamin D supplementation. Patients who met all of the eligibility criteria were randomized using a computer-generated system with block sizes of four to receive vitamin D (ergocalciferol 20,000 IU capsule, British Dispensary, Bangkok, Thailand) either high- or low-dose for 12 weeks.
The demographic data included gender, age, body mass index (BMI), type of hip fracture (intertrochanteric fracture or femoral neck fracture), treatment option either conservative or surgical procedures according to fracture type. The characteristics of previous functional status and activity (outdoor or indoor), and pre-operative ambulatory status whether patient could ambulate most time of the day independently without gait aid, depending on gait aids (single cane, tripod cane, quad cane, or pick up walker), or unable to walk were asked.
Patients in the low-dose vitamin D group received vitamin D2 20,000 IU (1 capsule) per week, while high-dose vitamin D group patients received 60,000 IU (3 capsules) per week. Generally, vitamin D was administered after surgical intervention or when the surgeon chose to treat a hip fracture conservatively. In addition, all patients received calcium carbonate 1,000 mg per day. Baseline serum 25(OH)D and calcium levels were measured prior to supplementation and at 12 weeks after the first dose of vitamin D. After the completion of this study, all patients in both treatment groups were switched to a maintenance dose of vitamin D2 supplementation (20,000 IU per week). The research assistants who collected data were blinded to each patient’s supplementation protocol.
Outcome measurement
The primary outcome of this study was the proportion of patients who achieved optimal vitamin D level at 12 weeks after vitamin D supplementation. The serum 25(OH)D concentration level was classified as hypovitaminosis D if the serum 25(OH)D level was < 30 ng/mL [11]. The quality assurance protocol of laboratory assays in this study was under the Thailand National External Quality Assessment Scheme (NEQAS) in clinical chemistry standard, which collaborated with the World Health Organization International External Quality Assessment Scheme (WHO IEQAS) [12, 13]. Serum 25(OH)D and parathyroid hormone (PTH) were measured and analyzed by electrochemiluminescence (ECL) binding assay on a cobas 8000 analyzer (Roche Diagnostics GmbH, Mannheim, Germany). The intra-assay coefficient of variation (CV) of serum 25(OH)D measurement at 10.6 and 29.0 ng/mL were 8.5% and 3.3%, respectively, while the inter-assay CV at 10.6 and 29.0 ng/mL.were 9.2% and 4.5%, respectively. For measurement of serum PTH level, an intra-assay CV at 49.3 and 160 pg/mL were 0.9% and 1.5%, respectively, while the inter-assay CV at 49.3 and 160 pg/mL were 0.8% and 1.3%, respectively. Total serum calcium was analyzed by 5-nitro-5’-methyl-BAPTA assay on a cobas 8000 analyzer (Roche Diagnostics GmbH, Mannheim, Germany), which the intra-assay CV at 9.3 and 14.6 mg/dL were 0.7% and 0.5%, respectively, while the inter-assay CV at 9.3 and 14.6 mg/dL were 0.8% and 0.9%, respectively. Hypercalcemia was classified as mild, moderate, or severe based on corrected serum calcium, which was calculated based on serum albumin level by adding 0.8 mg/dL of total serum calcium for every 1 mg decrease in serum albumin below 4 mg/dL, as described in the formula: corrected serum calcium (mg/dL) = measured total serum calcium (mg/dL) + [4.0 – serum albumin (g/dL) x 0.8] [14]. Mild hypercalcemia was defined as corrected serum calcium between 10.5 and 12 mg/dL, moderate hypercalcemia defined as corrected serum calcium from 12 to 14 mg/dL, and severe hypercalcemia was clarified when corrected serum calcium > 14 mg/dL [15]. Symptomatic hypercalcemia was diagnosed when corrected serum calcium was > 10.5 mg/dL with one of the following symptoms/signs: neurological dysfunction, myopathy, bradyarrhythmia, or gastrointestinal problems [16]. Functional outcome was evaluated using the Barthel Index and the EuroQol-visual analogue scale (EQ-VAS).
The Barthel Index is a measurement tool that is used to assess activities of daily living (ADLs). It comprises of 10 variables describing ADLs and mobility. Each item is rated by a scoring range which depended on performance of the patient. A higher score (maximum 100) indicates good functional ability and a greater likelihood of being able to live independently at home after discharge from the hospital, and a lower score (minimum 0) indicates poor functional status and likelihood of being dependent on caregiver. This tool has been validated in hip fracture patients [17].
EQ-VAS is a simple self-evaluated scale that rates health status ranging from 0 to 100 points. Patients rated their health status on a visual analogue scale, with a higher score (maximum 100) indicating the best imaginable health state, and a lower score (minimum 0) indicating the worst health status [18]. Similar to the Barthel Index, EQ-VAS was reported to be a reliable tool and has been validated in hip fracture patients [19].
Statistical analysis and sample size calculation
Statistical power was considered from the primary outcome, which was the proportion of patients with adequate level of 25(OH)D after supplementation for 12 weeks. A study by Sansanayudh N et al. [20] showed that among patients with metabolic syndrome who were supplemented with vitamin D2 20,000 IU per week for 8 weeks, 33.3% were restored to optimal vitamin D level, whereas 60% of patients who received vitamin D2 40,000 IU per week for 8 weeks were able to achieve optimal vitamin D level. Based on the results of that study, power analysis and sample size calculation indicated that a sample size of 54 patients per group would provide 80% statistical power (a one-sided alpha = 0.05; beta = 0.2). Recruitment was increased by 30% to compensate for a high incidence of loss to follow-up and death in this patient population. Therefore, a total of 140 patients were required for randomization in this study.
Baseline characteristics and all outcome measures are presented as number and percentage for categorial variables, and as mean ± standard deviation (SD) for continuous variables. The normality of data was assessed with Kolmogorov-Smirnov test. Pearson’s chi-square test or Fisher’s exact test was used for comparison of categorial variables, while Student’s t-test was used to compare continuous variables. Statistical analysis was performed using SPSS® Statistics version 18.0 (SPSS Inc., Chicago, IL, USA), and statistical significance was defined at a p-value of < 0.05.