Healthcare services utilization following admission for hip fracture in elderly patients

Healthcare services utilization following admission for hip fracture in elderly patients Abstract Objectives To assess the effect of hip fracture on healthcare utilization among elderly patients. Design Retrospective cohort study. Setting Eight general hospitals in Israel, owned by Clalit. Participants Enrollees >65 years, admitted with a hip fracture during 2009–2013. Main outcome measures Data collected included demographics, comorbidities, admission details related to the surgical and rehabilitation hospitalizations, mortality and costs. Mean monthly costs before and after the event were compared. Quantile regression was used to analyze associations between patient characteristics and healthcare expenditure in univariate and multivariate analysis. Results Of 9650 patients admitted with hip fracture during the study period, 6880 (71%) were Clalit enrollees and included in the present study (69% females, median age: 83 years). Total mean monthly costs increased by 96% during the follow-up year ($1470 vs. $749). Costs for rehabilitation accounted for 40% of costs during the first follow-up year. Mean monthly non-rehabilitation costs increased by 21% ($877 vs. $722). Several factors were found to be consistently associated with increased mean monthly costs during the follow-up year. These included Charlson’s comorbidity index, hypertension, baseline expenditure in the base year, the location of the fracture, procedure performed, department on admission, admission to the intensive care unit, discharge to a rehabilitation facility and mortality during the follow-up year. Conclusions Hip fractures in adults in Israel are associated with a significant increase in healthcare utilization and costs. The largest increment was seen in costs for rehabilitation. However, increased costs were noted in all sub-categories of healthcare costs. hip fracture, healthcare services utilization, public health, osteoporosis, health expenditure Introduction The incidence of hip fractures is rising, in parallel with the aging of the population and the prevalence of osteoporosis [1], and is estimated at 400–1 000 per 100 000 per year in the United States [2]. Performing hip fracture surgery within 48 h has been recommended by the OECD as a quality indicator [3]. A hip fracture significantly increases mortality to 20–26% within one year following the event [2, 4–8]. Mortality is six to eight times higher during the first 3 months following the fracture in females and males, respectively, and remains increased for 15 years [4]. Moreover, hip fractures were found to be a major cause of disability, resulting in 1.2 million years of life disabled worldwide [9] and a significant decrease in activity of daily living and instrumental activity of daily living scores. Therefore, most patients require rehabilitation following the acute treatment of the fracture [8]. Several studies assessed the impact of hip fracture on healthcare utilization and healthcare services costs, and demonstrated a significant increase in healthcare utilization in patients with hip fracture. In those studies, excess costs attributed to the fracture itself were estimated at $20 000–$37 000 per patient during the first year [10–12] These costs are mainly related to the acute hospital admission, but are also affected by treatment of long-term complications [11], and the functional disability that leads to repeated visits to rehabilitation centers and increased drug use, including pain medications, which might significantly contribute to drug costs [10]. Among patients with osteoporosis and a hip fracture, the event was associated with a 10-fold increase in healthcare-related costs during the year following the fracture, in comparison to a matched cohort without a fracture (4 014$ vs. 446$; P < 0.01). Total healthcare-related costs were almost doubled in the year following the fracture (15 942$ vs. 8 314$, P < 0.01) [10]. Most of these studies were conducted in the USA. However, expenditure and cost analyses are often difficult to generalize between countries, due to differences in healthcare systems and in actual treatment costs. In contrast with the private healthcare system in the USA, Israel has a public healthcare system, where pricing of healthcare-related costs is highly regulated by the Israeli Ministry of Health. In light of these major differences, we designed a study to evaluate these parameters in Israel, in patients enrolled in Clalit, the largest healthcare provider organizations in Israel, with 4.45 million enrollees. The main objective of this study was to assess the effect of hip fracture on healthcare utilization and costs in Clalit enrollees over the age of 65 years. This includes total monthly costs and cost-related sub-categories such as costs for rehabilitation, clinic visits, hospitalization for any reason and drug utilization. In addition, we examined other demographic and clinical characteristics and assessed their impact on healthcare utilization following an event of hip fracture. Methods and Materials The study population included Clalit enrollees hospitalized due to hip fracture (ICD-9-CM codes 820.XX) during 2009–2013 in any of eight general hospitals owned by Clalit. Data were extracted regarding demographics, underlying diagnoses, Charlson’s comorbidity index (CCI), lifestyle factors (alcohol abuse, smoking) and chronic medications. Additional data included the location of the fracture, the type and timing of surgery, the length of surgery, complications (blood loss, described in the operative report, transfer to the intensive care unit and reoperations during hospitalization), destination at discharge and mortality. The acute hospitalization due to hip fracture was defined as the index hospitalization. The ‘base year’ was defined as the 365 days prior to hospital admission and the ‘follow-up year’ was defined as the 365 days following hospital discharge. The primary outcome was the mean monthly costs in the follow-up year, as compared with the mean monthly costs in the base year. Secondary outcomes include sub-categories of healthcare utilization. Healthcare utilization data and the related costs during the base year were calculated and analyzed according to the following sub-categories: inpatient admissions, outpatient visits, emergency department visits, rehabilitation costs, laboratory costs, medication costs, day-care admissions and other costs. Costs in New Israeli Shekels (NIS) were converted to US$ using the mean annual conversion rate of 3.50 NIS per 1 US$ for 2017, as published by the Bank of Israel. Indirect costs were not calculated. Costs were standardized to the 2017 financial year terms using the healthcare expenditure component of the consumer price index, without discounting. We calculated and used mean monthly costs in order to compare costs during the ‘base year’ (for a full 12 months period) and the ‘follow-up year’ (for X months, depending on survival). This was performed because looking at annual expenditure could bias total costs in the hip fracture population, since we expected ~20% mortality during the first year following discharge. Statistical analysis Data were analyzed using SPSS for Windows, version 22 and Stata for Windows, version 12. Descriptive statistics were performed on all data variables. Nominal variables were described using distributions in percentages. Normally distributed continuous variables were described using mean and standard deviation and non-normally distributed variables were described using median and interquartile range. Identification of differences between the base year and follow-up year was performed using paired t-tests for normally distributed continuous variables. Univariate analysis was performed to identify associations between patient characteristics and healthcare utilization during the follow-up year. This was performed separately for patients with low, median and high healthcare utilization (25th, 50th and 75th percentile) using univariate quantile regression, which allows investigating various percentiles of the costs’ distribution and identifying differences in how the parameter influences various percentiles [13, 14] during the follow-up year in order to assess the impact of these variables on healthcare utilization. For the multivariate analysis of the primary outcome we again used quantile regression. Survival analysis was performed using the Kaplan–Meier method (for univariate analysis) while discrepancies between the groups were identified with the log rank test. The study was approved by the Institute Review Board of Meir Medical Center and by the Committee for Data Extraction at the General Management of Clalit. Patients’ informed consent was waived due to the retrospective nature of the study. Results Patient characteristics A total of 9650 patients >65 years were admitted to Clalit hospitals with hip fracture during the years 2009–2013, of which 6880 (71%) were eligible for the present study. The reasons for exclusion included non-Clalit enrollees (N = 1755), repeated fractures (N = 769), planned surgery for follow-up of fracture (N = 184), open fractures (N = 15), and trauma and motor vehicle accidents (N = 47). During the study period, the annual number of cases was relatively constant and ranged from 1340 (in 2011) to 1 405 (in 2010). The majority of the patients (69.1%) were females. Median age was 83 years. The most frequent comorbidities were hypertension (80.2%), hyperlipidemia (73.8%) and diabetes mellitus (35.4%). 21.5% of the patients had a previous stroke and 4.2% had a previous hip fracture (Table 1). Mean CCI score was 3.3 ± 2.6. The majority of the patients underwent surgery for the hip fracture during the index hospitalization (91.9%). Most of the fractures were either transcervical (35.3%) or trochanteric (38.6%). Most patients underwent closed reduction of the fracture with internal fixation (32.6%), partial hip replacement (30.0%) or Richard’s hip nailing procedure (27.0%). Two-thirds of the patients underwent surgery in <48 h of admission, while the mean time to surgery was 49.9 ± 54.3 h. About 11% required ICU hospitalization and 1.7% underwent re-operation during the hospital admission. Table 1 Patient characteristics (n = 6880) Variable  N (%) or mean ± SD  Female sex  4751 (69.1%)  Age (years)  Mean age  82.1 ± 7.3   65–74  1120 (16.3%)   75–84  2999 (43.6%)   85+  2761 (40.1%)  Locality   Urban Jewish  5629 (81.8%)   Rural Jewish  774 (11.2%)   Urban Arab  452 (6.6%)   Rural Arab  25 (0.4%)  Underlying diagnoses   Hypertension  5520 (80.2%)   Hyperlipidemia  5077 (73.8%)   Osteoporosis  2685 (39.0%)   Diabetes mellitus  2438 (35.4%)   Depression  1511 (22.0%)   History of stroke  1478 (21.5%)   Dementia  1429 (20.8%)   Hypothyroidism  1027 (14.9%)   Parkinson's disease  461 (6.7%)   Previous hip fracture  292 (4.2%)   Hyperthyroidism  217 (3.2%)   Epilepsy  161 (2.3%)   Rheumatoid arthritis  152 (2.2%)   Hypo/hyperparathyroidism  89 (1.3%)   Liver cirrhosis  87 (1.3%)  Charlson’s comorbidity index (n = 6871)  3.3 ± 2.6  Lifestyle factors   Alcohol use  82 (1.2%)   Smoking  1700 (24.7%)  Drugs   Proton-pump inhibitors  3219 (46.8%)   Selective serotonin reuptake inhibitors  1921 (27.9%)   Glucocorticosteroids  1635 (23.8%)   Antiepileptic drugs  1044 (15.2%)   Selective norepinephrine reuptake inhibitors  1027 (14.9%)   Vitamin K antagonists  847 (12.3%)   Heparin  503 (7.3%)  Underwent surgery  6326 (91.9%)  Fracture location   Intracapsular  18 (0.3%)   Midcervical  34 (0.5%)   Base  384 (5.6%)   Transcervical  2429 (35.3%)   Trochanteric  2656 (38.6%)   Intertrochanteric  765 (11.1%)   Sub-trochanteric  492 (6.9%)   Other/unspecified  119 (1.7%)  Surgery type   Closed reduction and internal fixation  2043 (32.6%)   Partial hip replacement  1880 (30.0%)   Richard’s hip nailing  1693 (27.0%)   Open reduction and internal fixation  445 (7.1%)   Total hip replacement  112 (1.8%)  Time to surgery <48 h  4118 (65.7%)  Complications   Blood loss  723 (11.5%)   Re-operation (at least one)  106 (1.7%)   Transfer to the intensive care unit  737 (10.7%)  Post-operative details  Discharge destination   Home  4871 (70.8%)   Rehabilitation  1809 (26.3%)   In hospital mortality  197 (2.9%)  Mortality at 1 year  1458 (21.2%)  Mortality at 5 years  3887 (56.5%)  Variable  N (%) or mean ± SD  Female sex  4751 (69.1%)  Age (years)  Mean age  82.1 ± 7.3   65–74  1120 (16.3%)   75–84  2999 (43.6%)   85+  2761 (40.1%)  Locality   Urban Jewish  5629 (81.8%)   Rural Jewish  774 (11.2%)   Urban Arab  452 (6.6%)   Rural Arab  25 (0.4%)  Underlying diagnoses   Hypertension  5520 (80.2%)   Hyperlipidemia  5077 (73.8%)   Osteoporosis  2685 (39.0%)   Diabetes mellitus  2438 (35.4%)   Depression  1511 (22.0%)   History of stroke  1478 (21.5%)   Dementia  1429 (20.8%)   Hypothyroidism  1027 (14.9%)   Parkinson's disease  461 (6.7%)   Previous hip fracture  292 (4.2%)   Hyperthyroidism  217 (3.2%)   Epilepsy  161 (2.3%)   Rheumatoid arthritis  152 (2.2%)   Hypo/hyperparathyroidism  89 (1.3%)   Liver cirrhosis  87 (1.3%)  Charlson’s comorbidity index (n = 6871)  3.3 ± 2.6  Lifestyle factors   Alcohol use  82 (1.2%)   Smoking  1700 (24.7%)  Drugs   Proton-pump inhibitors  3219 (46.8%)   Selective serotonin reuptake inhibitors  1921 (27.9%)   Glucocorticosteroids  1635 (23.8%)   Antiepileptic drugs  1044 (15.2%)   Selective norepinephrine reuptake inhibitors  1027 (14.9%)   Vitamin K antagonists  847 (12.3%)   Heparin  503 (7.3%)  Underwent surgery  6326 (91.9%)  Fracture location   Intracapsular  18 (0.3%)   Midcervical  34 (0.5%)   Base  384 (5.6%)   Transcervical  2429 (35.3%)   Trochanteric  2656 (38.6%)   Intertrochanteric  765 (11.1%)   Sub-trochanteric  492 (6.9%)   Other/unspecified  119 (1.7%)  Surgery type   Closed reduction and internal fixation  2043 (32.6%)   Partial hip replacement  1880 (30.0%)   Richard’s hip nailing  1693 (27.0%)   Open reduction and internal fixation  445 (7.1%)   Total hip replacement  112 (1.8%)  Time to surgery <48 h  4118 (65.7%)  Complications   Blood loss  723 (11.5%)   Re-operation (at least one)  106 (1.7%)   Transfer to the intensive care unit  737 (10.7%)  Post-operative details  Discharge destination   Home  4871 (70.8%)   Rehabilitation  1809 (26.3%)   In hospital mortality  197 (2.9%)  Mortality at 1 year  1458 (21.2%)  Mortality at 5 years  3887 (56.5%)  Post-operative and survival data The median length of stay (LOS) was 8 days. In-hospital mortality was 2.9%. Among patients who survived the index hospitalization, most were discharged home (70.8%), while 26.3% were discharged to rehabilitation. Kaplan–Meier estimates of survival were 78.8 and 43.5% at 1 and 5 years, respectively. Survival was better for females than for males (5-year survival: 47.9 vs. 33.6%, P < 0.001) and decreased with age (5-year survival: 63.1, 49.6 and 28.9% for age groups of 65–74, 75–84 and >85 years, respectively, P < 0.001). Healthcare utilization and healthcare-related costs Figure 1 demonstrates the changes in mean monthly healthcare-related costs in the base year and the follow-up year, in USD. Total mean monthly costs increased by 96% ($1 634 vs. $833, 5 719 vs. 2914 NIS, P < 0.001). Most of this increase was attributed to a 2 082% increase in mean monthly costs for rehabilitation ($660 vs. $30, 2 310 vs. 105 NIS, P < 0.001). Excluding rehabilitation costs, mean monthly costs were still 21% higher in the follow-up year ($975 vs. $803, 3 413 vs. 2 811 NIS, P < 0.001). This increase was attributed to an increase in inpatient costs (11% increase, $750 vs. $678, 2 624 vs. 2 373 NIS, P < 0.001), outpatient costs (160% increase, $134 vs. $52, 468 vs. 180 NIS, P < 0.001), day-care admissions (11% increase, $59 vs. $53, 205 vs. 184 NIS, P < 0.001), ED visits (27% increase, $12 vs. $9.4, 42 vs 33 NIS, P < 0.001), costs for laboratories (27% increase, $6.9 vs. $5.4, 24 vs. 19 NIS, P < 0.001) and medications costs (50% increase, $1.7 vs. $1.1, 6 vs. 4 NIS, P < 0.001, Fig. 1). Figure 1 View largeDownload slide Mean monthly healthcare-related costs (in 2017 USD) before and after hip fracture, according to healthcare sub-categories (1 USD = 3.5 New Israeli Shekels). Figure 1 View largeDownload slide Mean monthly healthcare-related costs (in 2017 USD) before and after hip fracture, according to healthcare sub-categories (1 USD = 3.5 New Israeli Shekels). Healthcare expenditure was higher for males; mean monthly costs in the base year were $960 (3361 NIS) and $776 (2715 NIS) for males and females, respectively (P < 0.001). During the follow-up year, a 107 and 91% increase in total mean monthly costs was seen for males and females, respectively (males: 1987$ [6955 NIS], females: 1482$ [5185 NIS], P < 0.001). The highest increase in costs (138%) was seen in patients aged over 85 years ($1 730 [6056 NIS] vs. $726 [2541 NIS], P < 0.001). Costs increased by 82 and 55% in the 75–84 age group (1562$ [5 469 NIS] vs. 858$ [3002 NIS]) and 65–74 age group (1 593$ [5576 NIS] vs. 1027$ [3593 NIS]), respectively. Healthcare expenditure was higher for patients who died during the follow-up year (mean monthly cost of $4861 [17 013 NIS], 380% increase from the baseline expenditure of 1 012$ [3541 NIS]). In patients who survived the follow-up year, total mean monthly costs increased by 13% ($890 [3.116 NIS] vs. 791$ [2769 NIS]). Factors associated with costs The univariate analysis for mean monthly costs in the follow-up year is presented in Table 2 Factors consistently associated with increased costs were male sex, higher CCI, liver cirrhosis, diabetes, hypertension, current smoking, history of stroke, medications (corticosteroids, progesterone, proton-pump inhibitors and vitamin K antagonists), baseline costs in the year before admission, the location of the fracture, length of surgery, time to surgery, LOSs in the orthopedics and intensive care unit, re-operation, costs of the index hospitalization and mortality (Table 2). Table 2 Univariate associations with healthcare-related costs during the follow-up year, in 2017 USD (1 USD = 3.5 New Israeli Shekels) Variable  Q1 (25th percentile)  Median (50th percentile)  Q3 (75th percentile)  Coefficient (95% CI)  P  Coefficient (95% CI)  P  Coefficient (95% CI)  P  Age (per year)  0.3 (−1.5, 2.0)  0.72  −1.2 (−3.9, 1.5)  0.38  1.8 (−8.9, 12.4)  0.75  Male  57.2 (29.1, 85.2)  <0.001  177.9 (132.8, 223.1)  <0.001  704.0 (544.4, 863.6)  <0.001  Charlson’s comorbidity index score (per point)  37.1 (28.4, 38.3)  <0.001  77.1 (68.7, 85.6)  <0.001  260.4 (238.3, 282.5)  <0.001  Cirrhosis  254.5 (141.3, 367.7)  <0.001  693.7 (509.1, 878.3)  <0.001  2 290 (1 573, 3 006)  <0.001  Diabetes  49.6 (21.5, 77.7)  0.001  122.0 (78.1, 166.0)  <0.001  558.5 (416.1, 701.0)  <0.001  Dementia  −128.8 (−164.9, −92.5)  <0.001  −126.1 (−173.6, −78.6)  <0.001  −125.2 (−302.6, 52.0)  0.17  Parkinson’s disease  39.9 (−8.4, 88.2)  0.11  107.9 (29.8, 186.1)  0.007  162.9 (−141.4, 467.4)  0.29  Hypertension  73.5 (38.7, 108.3)  <0.001  133.6 (84.5, 182.7)  <0.001  342.0 (146.8, 537.3)  0.001  Smoking  57.1 (26.1, 88.2)  <0.001  200.3 (153.3, 247.4)  <0.001  567.5 (407.4, 727.6)  <0.001  Previous stroke  67.7 (37.0, 98.3)  <0.001  158.4 (111.0, 205.9)  <0.001  500.3 (324.8, 675.8)  <0.001  Glucocorticosteroids  74.8 (43.6, 105.9)  <0.001  172.4 (122.7, 222.1)  <0.001  408.1 (237.9, 578.4)  <0.001  Heparin  49.1 (−0.4, 98.6)  0.052  221.3 (144.0, 298.5)  <0.001  854.9 (575.3, 1 134.5)  <0.001  Progesterone  233.0 (31.3, 434.7)  0.024  490.6 (172.6, 808.6)  0.002  2 079 (797.4, 3 360)  0.001  Proton-pump inhibitors  72.7 (46.7, 98.8)  <0.001  139.1 (96.4, 181.8)  <0.001  344.0 (186.9, 501.1)  <0.001  Vitamin K antagonists  135.6 (96.6, 174.8)  <0.001  275.3 (213.2, 337.3)  <0.001  863.6 (636.9, 1 090)  <0.001  Total mean monthly cost—base year (per 100 USD)  14.8 (13.2, 16.5)  <0.001  45.4 (42.9, 47.8)  <0.001  94.5 (87.3, 101.8)  <0.001  Sub-trochanteric fracture  60.5 (13.1, 107.7)  0.012  93.5 (16.1, 170.8)  0.018  112.5 (−206.7, 399.5)  0.53  Direct admission to orthopedics  −170.5 (−213.9, −127.2)  <0.001  −270.6 (−340.9, −200.4)  <0.001  −836.4 (−1 078, −594.3)  <0.001  Length of surgery (per hour)  62.1 (43.1, 81.1)  <0.001  89.0 (54.2, 123.9)  <0.001  192.0 (85.5, 298.6)  <0.001  Underwent surgery  177.4 (136.1, 218.7)  <0.001  −180.8 (−254.1, −107.5)  <0.001  −828.2 (−1 096, −560.5)  <0.001  Surgery within 48 h  36.2 (9.1, 63.3)  0.009  126.6 (82.7, 170.5)  <0.001  509.9 (349.1, 670.8)  <0.001  Time to surgery (per hour)  0.9 (0.6, 1.1)  <0.001  2.0 (1.6, 2.4)  <0.001  7.5 (6.0, 8.9)  <0.001  Total hip replacement  −173.4 (−268.1, −78.8)  <0.001  −240.8 (−383.0, −98.7)  0.001  −616.4 (−1 171, −62.5)  0.029  LOS (per day)  4.4 (2.8, 6.0)  <0.001  20.1 (18.0, 22.1)  <0.001  65.3 (56.3, 74.4)  <0.001  ICU admission  −195.2 (−237.6, −152.8)  <0.001  −10.2 (−71.6, 51.1)  0.74  6.0 (−241.4, 253.3)  0.96  Intensive care unit LOS (per day)  45.7 (30.3, 61.1)  <0.001  270.5 (243.3, 297.7)  <0.001  400.7 (187.1, 614.4)  <0.001  Orthopedics LOS (per day)  13.7 (10.9, 16.5)  <0.001  18.2 (14.2, 22.3)  <0.001  37.8 (22.0, 53.6)  <0.001  Re-operation  264.6 (193.4, 336.5)  <0.001  1 091 (964.0,1217)  <0.001  3304 (2 841, 3 769)  <0.001  Total admission cost (per 100 USD)  3.7 (3.4, 4.1)  <0.001  4.8 (4.3, 5.3)  <0.001  9.4 (7.4, 11.4)  <0.001  Discharge to rehabilitation  219.7 (187.7, 251.7)  <0.001  185.8 (133.6, 238.0)  <0.001  242.1 (66.5, 417.6)  0.007  Mortality at 1 year  1 176 (1 135, 1 216)  <0.001  3094 (3054.4, 3 134)  <0.001  5956 (5860, 6 052)  <0.001  Died during follow-up  130.4 (98.7, 162.1)  <0.001  487.8 (450.9, 524.7)  <0.001  2289 (2200, 2 378)  <0.001  Variable  Q1 (25th percentile)  Median (50th percentile)  Q3 (75th percentile)  Coefficient (95% CI)  P  Coefficient (95% CI)  P  Coefficient (95% CI)  P  Age (per year)  0.3 (−1.5, 2.0)  0.72  −1.2 (−3.9, 1.5)  0.38  1.8 (−8.9, 12.4)  0.75  Male  57.2 (29.1, 85.2)  <0.001  177.9 (132.8, 223.1)  <0.001  704.0 (544.4, 863.6)  <0.001  Charlson’s comorbidity index score (per point)  37.1 (28.4, 38.3)  <0.001  77.1 (68.7, 85.6)  <0.001  260.4 (238.3, 282.5)  <0.001  Cirrhosis  254.5 (141.3, 367.7)  <0.001  693.7 (509.1, 878.3)  <0.001  2 290 (1 573, 3 006)  <0.001  Diabetes  49.6 (21.5, 77.7)  0.001  122.0 (78.1, 166.0)  <0.001  558.5 (416.1, 701.0)  <0.001  Dementia  −128.8 (−164.9, −92.5)  <0.001  −126.1 (−173.6, −78.6)  <0.001  −125.2 (−302.6, 52.0)  0.17  Parkinson’s disease  39.9 (−8.4, 88.2)  0.11  107.9 (29.8, 186.1)  0.007  162.9 (−141.4, 467.4)  0.29  Hypertension  73.5 (38.7, 108.3)  <0.001  133.6 (84.5, 182.7)  <0.001  342.0 (146.8, 537.3)  0.001  Smoking  57.1 (26.1, 88.2)  <0.001  200.3 (153.3, 247.4)  <0.001  567.5 (407.4, 727.6)  <0.001  Previous stroke  67.7 (37.0, 98.3)  <0.001  158.4 (111.0, 205.9)  <0.001  500.3 (324.8, 675.8)  <0.001  Glucocorticosteroids  74.8 (43.6, 105.9)  <0.001  172.4 (122.7, 222.1)  <0.001  408.1 (237.9, 578.4)  <0.001  Heparin  49.1 (−0.4, 98.6)  0.052  221.3 (144.0, 298.5)  <0.001  854.9 (575.3, 1 134.5)  <0.001  Progesterone  233.0 (31.3, 434.7)  0.024  490.6 (172.6, 808.6)  0.002  2 079 (797.4, 3 360)  0.001  Proton-pump inhibitors  72.7 (46.7, 98.8)  <0.001  139.1 (96.4, 181.8)  <0.001  344.0 (186.9, 501.1)  <0.001  Vitamin K antagonists  135.6 (96.6, 174.8)  <0.001  275.3 (213.2, 337.3)  <0.001  863.6 (636.9, 1 090)  <0.001  Total mean monthly cost—base year (per 100 USD)  14.8 (13.2, 16.5)  <0.001  45.4 (42.9, 47.8)  <0.001  94.5 (87.3, 101.8)  <0.001  Sub-trochanteric fracture  60.5 (13.1, 107.7)  0.012  93.5 (16.1, 170.8)  0.018  112.5 (−206.7, 399.5)  0.53  Direct admission to orthopedics  −170.5 (−213.9, −127.2)  <0.001  −270.6 (−340.9, −200.4)  <0.001  −836.4 (−1 078, −594.3)  <0.001  Length of surgery (per hour)  62.1 (43.1, 81.1)  <0.001  89.0 (54.2, 123.9)  <0.001  192.0 (85.5, 298.6)  <0.001  Underwent surgery  177.4 (136.1, 218.7)  <0.001  −180.8 (−254.1, −107.5)  <0.001  −828.2 (−1 096, −560.5)  <0.001  Surgery within 48 h  36.2 (9.1, 63.3)  0.009  126.6 (82.7, 170.5)  <0.001  509.9 (349.1, 670.8)  <0.001  Time to surgery (per hour)  0.9 (0.6, 1.1)  <0.001  2.0 (1.6, 2.4)  <0.001  7.5 (6.0, 8.9)  <0.001  Total hip replacement  −173.4 (−268.1, −78.8)  <0.001  −240.8 (−383.0, −98.7)  0.001  −616.4 (−1 171, −62.5)  0.029  LOS (per day)  4.4 (2.8, 6.0)  <0.001  20.1 (18.0, 22.1)  <0.001  65.3 (56.3, 74.4)  <0.001  ICU admission  −195.2 (−237.6, −152.8)  <0.001  −10.2 (−71.6, 51.1)  0.74  6.0 (−241.4, 253.3)  0.96  Intensive care unit LOS (per day)  45.7 (30.3, 61.1)  <0.001  270.5 (243.3, 297.7)  <0.001  400.7 (187.1, 614.4)  <0.001  Orthopedics LOS (per day)  13.7 (10.9, 16.5)  <0.001  18.2 (14.2, 22.3)  <0.001  37.8 (22.0, 53.6)  <0.001  Re-operation  264.6 (193.4, 336.5)  <0.001  1 091 (964.0,1217)  <0.001  3304 (2 841, 3 769)  <0.001  Total admission cost (per 100 USD)  3.7 (3.4, 4.1)  <0.001  4.8 (4.3, 5.3)  <0.001  9.4 (7.4, 11.4)  <0.001  Discharge to rehabilitation  219.7 (187.7, 251.7)  <0.001  185.8 (133.6, 238.0)  <0.001  242.1 (66.5, 417.6)  0.007  Mortality at 1 year  1 176 (1 135, 1 216)  <0.001  3094 (3054.4, 3 134)  <0.001  5956 (5860, 6 052)  <0.001  Died during follow-up  130.4 (98.7, 162.1)  <0.001  487.8 (450.9, 524.7)  <0.001  2289 (2200, 2 378)  <0.001  ICU, intensive care unit; LOS, length of stay; USD, United States dollars. A multivariate quantile regression model for mean monthly costs in the follow-up year is presented in Table 3. Factors consistently associated with increased mean monthly costs were CCI, hypertension, baseline expenditure in the base year, the location of the fracture, the hospital LOS, discharge to a rehabilitation facility and mortality (Table 3). Table 3 Multivariate quantile regression model for mean monthly costs in the follow-up year, in 2017 USD (1 USD = 3.5 New Israeli Shekels) Variable  Q1 (25th percentile)  Median (50th percentile)  Q3 (75th percentile)  Coefficient (95% CI)  P  Coefficient (95% CI)  P  Coefficient (95% CI)  P  Age (per year)  −2.1 (−4.0, −0.2)  0.032  −1.9 (−4.3, 0.6)  0.134  −5.7 (−10.1, −1.2)  0.012  Male  8.2 (−20.1, 36.6)  0.570  38.1 (−0.1, 76.3)  0.051  108.9 (40.4, 177.4)  0.002  Charlson comorbidity index score (per point)  22.1 (16.7, 27.4)  <0.001  23.5 (16.1, 30.7)  <0.001  55.1 (41.7, 68.7)  <0.001  Cirrhosis  204.9 (82.9, 327.1)  0.001  247.1 (81.6, 412.5)  0.003  –    Dementia  −161.9 (−193.8, −130.0)  <0.001  −196.9 (−240.7, −153.0)  <0.001  −207.1 (−286.1, −128.0)  <0.001  Hypertension  65.7 (32.9, 98.5)  <0.001  63.5 (19.2, 107.7)  0.005  106.5 (26.3, 186.5)  0.009  Parkinson’s disease  –    95.5 (26.9, 163.9)  0.006  –    Vitamin K antagonists  45.3 (5.4, 85.2)  0.026  74.8 (20.8, 128.8)  0.007  –    Total mean monthly cost—base year (per 100 SD)  2.9 (2.4, 3.4)  <0.001  26.7 (24.7, 28.8)  <0.001  53.7 (49.8, 57.6)  <0.001  Sub-trochanteric fracture  64.7 (15.6, 113.8)  0.010  113.6 (47.1, 179.9)  0.001  154.4 (34.3, 274.4)  0.012  Underwent surgery  −406.3 (−692.0, −120.7)  0.005  –    –    Total hip replacement  −101.8 (−197.2, −6.4)  0.036  −226.1 (−355.7, −96.5)  0.001  −369.4 (−593.5, −123.1)  0.003  Direct admission to Orthopedics  −359.1 (−420.2, −297.9)  <0.001  −691.3 (−765.7, −616.8)  <0.001  −1 131.9 (−1 256.1, −1 007.6)  <0.001  LOS (per day)  3.0 (1.2, 4.7)  0.001  15.6 (13.2, 17.9)  <0.001  33.3 (29.0, 37.7)  <0.001  ICU admission  −246.0 (−304.8, −187.0)  <0.001  −490.1 (−564.0, −416.3)  <0.001  −754.0 (−881.7, −626.3)  <0.001  Discharge to rehabilitation  212.7 (183.9, 241.6)  <0.001  143.4 (104.4, 182.5)  <0.001  191.6 (120.8, 262.5)  <0.001  1-year mortality  1 203.7 (1 169, 1 239)  <0.001  3 044.2 (2 996, 3 091)  <0.001  5 531.6 (5 448, 5 615)  <0.001  Constant  1 002    950.2    1 632    Variable  Q1 (25th percentile)  Median (50th percentile)  Q3 (75th percentile)  Coefficient (95% CI)  P  Coefficient (95% CI)  P  Coefficient (95% CI)  P  Age (per year)  −2.1 (−4.0, −0.2)  0.032  −1.9 (−4.3, 0.6)  0.134  −5.7 (−10.1, −1.2)  0.012  Male  8.2 (−20.1, 36.6)  0.570  38.1 (−0.1, 76.3)  0.051  108.9 (40.4, 177.4)  0.002  Charlson comorbidity index score (per point)  22.1 (16.7, 27.4)  <0.001  23.5 (16.1, 30.7)  <0.001  55.1 (41.7, 68.7)  <0.001  Cirrhosis  204.9 (82.9, 327.1)  0.001  247.1 (81.6, 412.5)  0.003  –    Dementia  −161.9 (−193.8, −130.0)  <0.001  −196.9 (−240.7, −153.0)  <0.001  −207.1 (−286.1, −128.0)  <0.001  Hypertension  65.7 (32.9, 98.5)  <0.001  63.5 (19.2, 107.7)  0.005  106.5 (26.3, 186.5)  0.009  Parkinson’s disease  –    95.5 (26.9, 163.9)  0.006  –    Vitamin K antagonists  45.3 (5.4, 85.2)  0.026  74.8 (20.8, 128.8)  0.007  –    Total mean monthly cost—base year (per 100 SD)  2.9 (2.4, 3.4)  <0.001  26.7 (24.7, 28.8)  <0.001  53.7 (49.8, 57.6)  <0.001  Sub-trochanteric fracture  64.7 (15.6, 113.8)  0.010  113.6 (47.1, 179.9)  0.001  154.4 (34.3, 274.4)  0.012  Underwent surgery  −406.3 (−692.0, −120.7)  0.005  –    –    Total hip replacement  −101.8 (−197.2, −6.4)  0.036  −226.1 (−355.7, −96.5)  0.001  −369.4 (−593.5, −123.1)  0.003  Direct admission to Orthopedics  −359.1 (−420.2, −297.9)  <0.001  −691.3 (−765.7, −616.8)  <0.001  −1 131.9 (−1 256.1, −1 007.6)  <0.001  LOS (per day)  3.0 (1.2, 4.7)  0.001  15.6 (13.2, 17.9)  <0.001  33.3 (29.0, 37.7)  <0.001  ICU admission  −246.0 (−304.8, −187.0)  <0.001  −490.1 (−564.0, −416.3)  <0.001  −754.0 (−881.7, −626.3)  <0.001  Discharge to rehabilitation  212.7 (183.9, 241.6)  <0.001  143.4 (104.4, 182.5)  <0.001  191.6 (120.8, 262.5)  <0.001  1-year mortality  1 203.7 (1 169, 1 239)  <0.001  3 044.2 (2 996, 3 091)  <0.001  5 531.6 (5 448, 5 615)  <0.001  Constant  1 002    950.2    1 632    ICU, intensive care unit; LOS, length of stay; USD, United States dollars. Discussion Our results demonstrate that hip fracture in the elderly population in Israel is associated with significantly increased healthcare utilization and costs. During the year following a hip fracture, mean monthly costs increased by 96% in comparison to the year before the fracture. Rehabilitation costs comprised of ~40% of the total mean monthly costs during the follow-up year. The increments in healthcare expenditures were higher in males as well as in patients who died during the follow-up year, and increased with age. Multivariate regression identified factors associated with increased mean monthly costs during the follow-up year, which included higher CCI score, certain comorbidities, expenditure in the base year, discharge to a rehabilitation facility and mortality. These data correspond with previous studies [2, 5, 7, 8, 15]. A recent study which examined 6019 patients with hip fracture and assessed costs before and after hospital admission [11], showed that healthcare expenditure increased by 147% during the year following the fracture. Of these costs, 64% were attributed to the acute care of the fracture and 36% to rehabilitation [11]. Similarly, a Canadian study that compared mean annual costs between patients with hip fracture and a matched cohort showed that costs in the hip fracture group increased by 241 and 266% in females and males, respectively [12]. The largest component of costs attributable to the hip fracture was acute hospitalizations, accounting for ~40% (38–41%). Similar to our study, higher costs were demonstrated in patients who died during follow-up. Another recent study analyzed healthcare utilization and costs in 8028 patients with hip fracture in the UK [16]. Mean annual costs increased by ~135%. Inpatient costs comprised 92% of the total costs during the follow-up year (vs. 83% in the base year). This study also demonstrated that variables associated with significantly higher healthcare costs in the follow-up period after hospitalization included age >65 years, higher CCI score, discharge to another institution, hospitalizations and outpatients visits in the baseline year and male sex [16]. Another American study [17] demonstrated a 400% increase in mean healthcare expenditures in the year following hip fracture. Approximately 88% of this increment has been directly related to the hip fracture. Healthcare expenditures increased in all sub-categories of healthcare utilization, except cancer treatment-related costs. These include costs for treatment directly related to the fracture (aftercare, complications, decubitus ulcers) and costs for treatment unrelated to the fracture (cardiovascular, respiratory) [17]. Although differences exist in study designs and setting, all these studies show similar trends: healthcare utilization is at least doubled in the year following a hip fracture. Indeed, in the present study, costs for rehabilitation accounted for <50% of the total costs during the follow-up year. Some of the economic effects demonstrated in this study are quite expected. Hip fracture requires rehabilitation in the majority of cases. Functional health status limitations were previously [5] found in all ADL categories among patients with hip fracture in comparison to controls. Thus, it is no surprise that costs for rehabilitation increased dramatically in the present study. However, we have shown that mean monthly costs have increased in all sub-categories of healthcare utilization during the follow-up year, with a 21% increase in non-rehabilitation costs. A question arises regarding how much of this increase in costs is attributed to the fracture itself. Following the hip fracture, 51% of the patients were found to be physically frail and 34% were disabled (defined as any ADL dependency) [18]. Some degree of cognitive impairment was found in 89% of the patients. Complications were seen in 71% of patients with frailty and cognitive impairment [18]. It was recently shown that patients with frailty utilize healthcare services to a greater extent in comparison to matched patients without frailty [19]. This includes increased general practitioner and outpatient visits, hospitalizations and nursing care. Assuming that a certain proportion of patients have physical and/or cognitive frailty following a hip fracture, it is possible that frailty accounts for a portion of the increased costs during the following year. Therefore, frailty should be a target for rehabilitation, monitoring and treatment, both from the patient safety and quality of life perspective, and from the economic perspective. Another factor to be considered is aging. According to Clalit HMO’s database, healthcare utilization in the general population increases in ~2% per year, due to aging alone. Therefore, the much higher increase in costs seen in the present study is unlikely to be explained by aging alone. Patients who died during the follow-up year had the largest increase in healthcare-related costs. One can assume that they were in a poorer post- and maybe pre-fracture medical condition, and probably required more intensive medical care following the fracture. The present study has some strengths, including a large sample size, accurate billing information collected directly from an administrative database, complete follow-up for all Clalit enrollees included in the study, both in the hospital and in the community settings, and a unique opportunity to explore this topic in a public healthcare system with government dictated healthcare services coverage. Our study has several limitations. Costs were calculated according to Clalit’s database, which includes only direct costs. These costs are either calculated based on the LOS with a per-diem cost, or based on a prospective payment diagnosis-related group-like mechanism, where several procedure codes are compensated on a global basis. However, this method does not take into account detailed costs of procedures (e.g. magnetic resonance imaging) or costly medications (e.g. broad-spectrum antibiotics) given during hospital admission. In addition, indirect costs, such as loss of work days, home assistance that may be required for some patients, and supportive medical devices not covered by the HMO, were not taken into account in our calculations. No quality of life data were obtained, thus the impact of hip fracture on this parameter and interrelations between it and healthcare utilization were not assessed. Some Clalit hospitals have a rehabilitation facility within them. Thus, in some hospitals, costs for rehabilitation (at least partially) were included in the costs for the index hospitalization. This means that at least some of the costs for the index hospitalization are actually attributed to rehabilitation. The study included only Clalit enrollees; however, since most of the basic medical coverage in Israel is dictated by the Ministry of Health, we can assume that costs for these services apply to the other three HMOs in Israel, as well. Finally, some demographic differences between Clalit enrollees and enrollees in the other three HMOs cannot be ruled out. To conclude, our study demonstrates that hip fracture in patients over 65 years of age is associated with a significant increase in total mean monthly costs during the follow-up year, for all examined sub-categories of costs. As the population worldwide is aging and life expectancy increases, it is reasonable to assume that the incidence of hip fractures, and the economic burden it brings, will only increase in the future. Additional study is therefore required to further characterize this increase in costs and to find and define modifiable factors that may affect these changes in healthcare utilization following a hip fracture. Acknowledgements This study was conducted as part of the requirements for graduation from the medical school of the Faculty of Health Sciences, Ben-Gurion University of the Negev. The study was presented at the 33rd conference of the International Society for Quality in Healthcare, Tokyo 2016. References 1 Nayagam S. Injuries of the hip and femur. In: Solomon L, Waraick D, Nayagam S (eds). Apley’s System of Orthopaedics and Fractures , 9th edn. Bristol, UK: Hodder Arnold, 2010: 843– 74. Google Scholar CrossRef Search ADS   2 Brauer CA, Coca-Perraillon M, Cutler DM et al.  . Incidence and mortality of hip fractures in the United States. J Am Med Assoc  2009; 302: 1573– 9. Google Scholar CrossRef Search ADS   3 Carinci F, Van Gool K, Mainz J et al.  . Towards actionable international comparisons of health system performance: expert revision of the OECD framework and quality indicators. Int J Qual Health Care  2015; 27: 137– 46. Google Scholar PubMed  4 Haentjens P, Magaziner J, Colon-Emeric CS et al.  . Meta-analysis: excess mortality after hip fracture among older women and men. Ann Intern Med  2010; 152: 380– 90. Google Scholar CrossRef Search ADS PubMed  5 Wolinsky FD, Fitzgerald JF, Stump TE. The effect of hip fracture on mortality, hospitalization, and functional status: a prospective study. Am J Public Health  1997; 87: 398– 403. Google Scholar CrossRef Search ADS PubMed  6 LeBlanc ES, Hillier TA, Pedula KL et al.  . Hip fracture and increased short-term but not long-term mortality in healthy older women. Arch Intern Med  2011; 170: 1831– 7. Google Scholar CrossRef Search ADS   7 Tosteson ANA, Gottlieb DJ, Radley D et al.  . Excess mortality following hip fracture: the role of underlying health status. Osteoporosis Int  2007; 18: 1463– 72. Google Scholar CrossRef Search ADS   8 Bentler SE, Liu L, Obrizan M et al.  . The aftermath of hip fracture: discharge placement, functional status change, and mortality. Am J Epidemiol  2009; 170: 1290– 9. Google Scholar CrossRef Search ADS PubMed  9 Johnell O, Kanis JA. An estimate of the worldwide prevalence, mortality and disability associated with hip fracture. Osteoporosis Int  2004; 15: 897– 902. Google Scholar CrossRef Search ADS   10 Orsini LS, Rousculp MD, Long SR et al.  . Health care utilization and expenditures in the United States: a study of osteoporosis-related fractures. Osteoporosis Int  2005; 16: 359– 71. Google Scholar CrossRef Search ADS   11 Duclos A, Souray-Targe S, Randrianasolo M et al.  . Burden of hip fracture on inpatient care: a before and after population-based study. Osteoporosis Int  2010; 21: 1493– 1501. Google Scholar CrossRef Search ADS   12 Nikitovic M, Wodchis WP, Krahn MD et al.  . Direct health-care costs attributed to hip fractures among seniors: a matched cohort study. Osteoporosis Int  2013; 24: 659– 69. Google Scholar CrossRef Search ADS   13 Grebregziagher M, Lynch CP, Mueller M et al.  . Using quantile regression to investigate racial disparities in medication non-adherence. BMC Med Res Methodol  2011; 11: 88. Google Scholar CrossRef Search ADS PubMed  14 Rehkopf DH. Quantile regression for hypothesis testing and hypothesis screening at the dawn of big data. Epidemiology  2012; 23: 665– 7. Google Scholar CrossRef Search ADS PubMed  15 Panula J, Pihlajamaki H, Mattila VM et al.  . Mortality and cause of death in hip fracture patients aged 65 or older—a population-based study. BMC Musculoskelet Disord  2011; 12: 105. Google Scholar CrossRef Search ADS PubMed  16 Lambrelli D, Burge R, Raluy-Callado M et al.  . Retrospective database study to assess the economic impact of hip fracture in the United Kingdom. J Med Econ  2014; 17: 817– 25. Google Scholar CrossRef Search ADS PubMed  17 Kilgore ML, Curtis JR, Delzell E et al.  . A close examination of healthcare expenditures related to fractures. J Bone Mineral Res  2013; 28: 816– 20. Google Scholar CrossRef Search ADS   18 Kistler EA, Nicholas JA, Kates SL et al.  . Frailty and short-term outcomes in patients with hip fracture. Ger Ortho Surg Rehab  2015; 6: 209– 14. Google Scholar CrossRef Search ADS   19 Coelho T, Paul C, Gobbens RJJ et al.  . Frailty as a predictor of short-term adverse outcomes. PeerJ  2015; 3: e1121. Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2017. Published by Oxford University Press in association with the International Society for Quality in Health Care. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png International Journal for Quality in Health Care Oxford University Press

Healthcare services utilization following admission for hip fracture in elderly patients

Loading next page...
 
/lp/ou_press/healthcare-services-utilization-following-admission-for-hip-fracture-XmzXX0wKQe
Publisher
Oxford University Press
Copyright
© The Author(s) 2017. Published by Oxford University Press in association with the International Society for Quality in Health Care. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com
ISSN
1353-4505
eISSN
1464-3677
D.O.I.
10.1093/intqhc/mzx178
Publisher site
See Article on Publisher Site

Abstract

Abstract Objectives To assess the effect of hip fracture on healthcare utilization among elderly patients. Design Retrospective cohort study. Setting Eight general hospitals in Israel, owned by Clalit. Participants Enrollees >65 years, admitted with a hip fracture during 2009–2013. Main outcome measures Data collected included demographics, comorbidities, admission details related to the surgical and rehabilitation hospitalizations, mortality and costs. Mean monthly costs before and after the event were compared. Quantile regression was used to analyze associations between patient characteristics and healthcare expenditure in univariate and multivariate analysis. Results Of 9650 patients admitted with hip fracture during the study period, 6880 (71%) were Clalit enrollees and included in the present study (69% females, median age: 83 years). Total mean monthly costs increased by 96% during the follow-up year ($1470 vs. $749). Costs for rehabilitation accounted for 40% of costs during the first follow-up year. Mean monthly non-rehabilitation costs increased by 21% ($877 vs. $722). Several factors were found to be consistently associated with increased mean monthly costs during the follow-up year. These included Charlson’s comorbidity index, hypertension, baseline expenditure in the base year, the location of the fracture, procedure performed, department on admission, admission to the intensive care unit, discharge to a rehabilitation facility and mortality during the follow-up year. Conclusions Hip fractures in adults in Israel are associated with a significant increase in healthcare utilization and costs. The largest increment was seen in costs for rehabilitation. However, increased costs were noted in all sub-categories of healthcare costs. hip fracture, healthcare services utilization, public health, osteoporosis, health expenditure Introduction The incidence of hip fractures is rising, in parallel with the aging of the population and the prevalence of osteoporosis [1], and is estimated at 400–1 000 per 100 000 per year in the United States [2]. Performing hip fracture surgery within 48 h has been recommended by the OECD as a quality indicator [3]. A hip fracture significantly increases mortality to 20–26% within one year following the event [2, 4–8]. Mortality is six to eight times higher during the first 3 months following the fracture in females and males, respectively, and remains increased for 15 years [4]. Moreover, hip fractures were found to be a major cause of disability, resulting in 1.2 million years of life disabled worldwide [9] and a significant decrease in activity of daily living and instrumental activity of daily living scores. Therefore, most patients require rehabilitation following the acute treatment of the fracture [8]. Several studies assessed the impact of hip fracture on healthcare utilization and healthcare services costs, and demonstrated a significant increase in healthcare utilization in patients with hip fracture. In those studies, excess costs attributed to the fracture itself were estimated at $20 000–$37 000 per patient during the first year [10–12] These costs are mainly related to the acute hospital admission, but are also affected by treatment of long-term complications [11], and the functional disability that leads to repeated visits to rehabilitation centers and increased drug use, including pain medications, which might significantly contribute to drug costs [10]. Among patients with osteoporosis and a hip fracture, the event was associated with a 10-fold increase in healthcare-related costs during the year following the fracture, in comparison to a matched cohort without a fracture (4 014$ vs. 446$; P < 0.01). Total healthcare-related costs were almost doubled in the year following the fracture (15 942$ vs. 8 314$, P < 0.01) [10]. Most of these studies were conducted in the USA. However, expenditure and cost analyses are often difficult to generalize between countries, due to differences in healthcare systems and in actual treatment costs. In contrast with the private healthcare system in the USA, Israel has a public healthcare system, where pricing of healthcare-related costs is highly regulated by the Israeli Ministry of Health. In light of these major differences, we designed a study to evaluate these parameters in Israel, in patients enrolled in Clalit, the largest healthcare provider organizations in Israel, with 4.45 million enrollees. The main objective of this study was to assess the effect of hip fracture on healthcare utilization and costs in Clalit enrollees over the age of 65 years. This includes total monthly costs and cost-related sub-categories such as costs for rehabilitation, clinic visits, hospitalization for any reason and drug utilization. In addition, we examined other demographic and clinical characteristics and assessed their impact on healthcare utilization following an event of hip fracture. Methods and Materials The study population included Clalit enrollees hospitalized due to hip fracture (ICD-9-CM codes 820.XX) during 2009–2013 in any of eight general hospitals owned by Clalit. Data were extracted regarding demographics, underlying diagnoses, Charlson’s comorbidity index (CCI), lifestyle factors (alcohol abuse, smoking) and chronic medications. Additional data included the location of the fracture, the type and timing of surgery, the length of surgery, complications (blood loss, described in the operative report, transfer to the intensive care unit and reoperations during hospitalization), destination at discharge and mortality. The acute hospitalization due to hip fracture was defined as the index hospitalization. The ‘base year’ was defined as the 365 days prior to hospital admission and the ‘follow-up year’ was defined as the 365 days following hospital discharge. The primary outcome was the mean monthly costs in the follow-up year, as compared with the mean monthly costs in the base year. Secondary outcomes include sub-categories of healthcare utilization. Healthcare utilization data and the related costs during the base year were calculated and analyzed according to the following sub-categories: inpatient admissions, outpatient visits, emergency department visits, rehabilitation costs, laboratory costs, medication costs, day-care admissions and other costs. Costs in New Israeli Shekels (NIS) were converted to US$ using the mean annual conversion rate of 3.50 NIS per 1 US$ for 2017, as published by the Bank of Israel. Indirect costs were not calculated. Costs were standardized to the 2017 financial year terms using the healthcare expenditure component of the consumer price index, without discounting. We calculated and used mean monthly costs in order to compare costs during the ‘base year’ (for a full 12 months period) and the ‘follow-up year’ (for X months, depending on survival). This was performed because looking at annual expenditure could bias total costs in the hip fracture population, since we expected ~20% mortality during the first year following discharge. Statistical analysis Data were analyzed using SPSS for Windows, version 22 and Stata for Windows, version 12. Descriptive statistics were performed on all data variables. Nominal variables were described using distributions in percentages. Normally distributed continuous variables were described using mean and standard deviation and non-normally distributed variables were described using median and interquartile range. Identification of differences between the base year and follow-up year was performed using paired t-tests for normally distributed continuous variables. Univariate analysis was performed to identify associations between patient characteristics and healthcare utilization during the follow-up year. This was performed separately for patients with low, median and high healthcare utilization (25th, 50th and 75th percentile) using univariate quantile regression, which allows investigating various percentiles of the costs’ distribution and identifying differences in how the parameter influences various percentiles [13, 14] during the follow-up year in order to assess the impact of these variables on healthcare utilization. For the multivariate analysis of the primary outcome we again used quantile regression. Survival analysis was performed using the Kaplan–Meier method (for univariate analysis) while discrepancies between the groups were identified with the log rank test. The study was approved by the Institute Review Board of Meir Medical Center and by the Committee for Data Extraction at the General Management of Clalit. Patients’ informed consent was waived due to the retrospective nature of the study. Results Patient characteristics A total of 9650 patients >65 years were admitted to Clalit hospitals with hip fracture during the years 2009–2013, of which 6880 (71%) were eligible for the present study. The reasons for exclusion included non-Clalit enrollees (N = 1755), repeated fractures (N = 769), planned surgery for follow-up of fracture (N = 184), open fractures (N = 15), and trauma and motor vehicle accidents (N = 47). During the study period, the annual number of cases was relatively constant and ranged from 1340 (in 2011) to 1 405 (in 2010). The majority of the patients (69.1%) were females. Median age was 83 years. The most frequent comorbidities were hypertension (80.2%), hyperlipidemia (73.8%) and diabetes mellitus (35.4%). 21.5% of the patients had a previous stroke and 4.2% had a previous hip fracture (Table 1). Mean CCI score was 3.3 ± 2.6. The majority of the patients underwent surgery for the hip fracture during the index hospitalization (91.9%). Most of the fractures were either transcervical (35.3%) or trochanteric (38.6%). Most patients underwent closed reduction of the fracture with internal fixation (32.6%), partial hip replacement (30.0%) or Richard’s hip nailing procedure (27.0%). Two-thirds of the patients underwent surgery in <48 h of admission, while the mean time to surgery was 49.9 ± 54.3 h. About 11% required ICU hospitalization and 1.7% underwent re-operation during the hospital admission. Table 1 Patient characteristics (n = 6880) Variable  N (%) or mean ± SD  Female sex  4751 (69.1%)  Age (years)  Mean age  82.1 ± 7.3   65–74  1120 (16.3%)   75–84  2999 (43.6%)   85+  2761 (40.1%)  Locality   Urban Jewish  5629 (81.8%)   Rural Jewish  774 (11.2%)   Urban Arab  452 (6.6%)   Rural Arab  25 (0.4%)  Underlying diagnoses   Hypertension  5520 (80.2%)   Hyperlipidemia  5077 (73.8%)   Osteoporosis  2685 (39.0%)   Diabetes mellitus  2438 (35.4%)   Depression  1511 (22.0%)   History of stroke  1478 (21.5%)   Dementia  1429 (20.8%)   Hypothyroidism  1027 (14.9%)   Parkinson's disease  461 (6.7%)   Previous hip fracture  292 (4.2%)   Hyperthyroidism  217 (3.2%)   Epilepsy  161 (2.3%)   Rheumatoid arthritis  152 (2.2%)   Hypo/hyperparathyroidism  89 (1.3%)   Liver cirrhosis  87 (1.3%)  Charlson’s comorbidity index (n = 6871)  3.3 ± 2.6  Lifestyle factors   Alcohol use  82 (1.2%)   Smoking  1700 (24.7%)  Drugs   Proton-pump inhibitors  3219 (46.8%)   Selective serotonin reuptake inhibitors  1921 (27.9%)   Glucocorticosteroids  1635 (23.8%)   Antiepileptic drugs  1044 (15.2%)   Selective norepinephrine reuptake inhibitors  1027 (14.9%)   Vitamin K antagonists  847 (12.3%)   Heparin  503 (7.3%)  Underwent surgery  6326 (91.9%)  Fracture location   Intracapsular  18 (0.3%)   Midcervical  34 (0.5%)   Base  384 (5.6%)   Transcervical  2429 (35.3%)   Trochanteric  2656 (38.6%)   Intertrochanteric  765 (11.1%)   Sub-trochanteric  492 (6.9%)   Other/unspecified  119 (1.7%)  Surgery type   Closed reduction and internal fixation  2043 (32.6%)   Partial hip replacement  1880 (30.0%)   Richard’s hip nailing  1693 (27.0%)   Open reduction and internal fixation  445 (7.1%)   Total hip replacement  112 (1.8%)  Time to surgery <48 h  4118 (65.7%)  Complications   Blood loss  723 (11.5%)   Re-operation (at least one)  106 (1.7%)   Transfer to the intensive care unit  737 (10.7%)  Post-operative details  Discharge destination   Home  4871 (70.8%)   Rehabilitation  1809 (26.3%)   In hospital mortality  197 (2.9%)  Mortality at 1 year  1458 (21.2%)  Mortality at 5 years  3887 (56.5%)  Variable  N (%) or mean ± SD  Female sex  4751 (69.1%)  Age (years)  Mean age  82.1 ± 7.3   65–74  1120 (16.3%)   75–84  2999 (43.6%)   85+  2761 (40.1%)  Locality   Urban Jewish  5629 (81.8%)   Rural Jewish  774 (11.2%)   Urban Arab  452 (6.6%)   Rural Arab  25 (0.4%)  Underlying diagnoses   Hypertension  5520 (80.2%)   Hyperlipidemia  5077 (73.8%)   Osteoporosis  2685 (39.0%)   Diabetes mellitus  2438 (35.4%)   Depression  1511 (22.0%)   History of stroke  1478 (21.5%)   Dementia  1429 (20.8%)   Hypothyroidism  1027 (14.9%)   Parkinson's disease  461 (6.7%)   Previous hip fracture  292 (4.2%)   Hyperthyroidism  217 (3.2%)   Epilepsy  161 (2.3%)   Rheumatoid arthritis  152 (2.2%)   Hypo/hyperparathyroidism  89 (1.3%)   Liver cirrhosis  87 (1.3%)  Charlson’s comorbidity index (n = 6871)  3.3 ± 2.6  Lifestyle factors   Alcohol use  82 (1.2%)   Smoking  1700 (24.7%)  Drugs   Proton-pump inhibitors  3219 (46.8%)   Selective serotonin reuptake inhibitors  1921 (27.9%)   Glucocorticosteroids  1635 (23.8%)   Antiepileptic drugs  1044 (15.2%)   Selective norepinephrine reuptake inhibitors  1027 (14.9%)   Vitamin K antagonists  847 (12.3%)   Heparin  503 (7.3%)  Underwent surgery  6326 (91.9%)  Fracture location   Intracapsular  18 (0.3%)   Midcervical  34 (0.5%)   Base  384 (5.6%)   Transcervical  2429 (35.3%)   Trochanteric  2656 (38.6%)   Intertrochanteric  765 (11.1%)   Sub-trochanteric  492 (6.9%)   Other/unspecified  119 (1.7%)  Surgery type   Closed reduction and internal fixation  2043 (32.6%)   Partial hip replacement  1880 (30.0%)   Richard’s hip nailing  1693 (27.0%)   Open reduction and internal fixation  445 (7.1%)   Total hip replacement  112 (1.8%)  Time to surgery <48 h  4118 (65.7%)  Complications   Blood loss  723 (11.5%)   Re-operation (at least one)  106 (1.7%)   Transfer to the intensive care unit  737 (10.7%)  Post-operative details  Discharge destination   Home  4871 (70.8%)   Rehabilitation  1809 (26.3%)   In hospital mortality  197 (2.9%)  Mortality at 1 year  1458 (21.2%)  Mortality at 5 years  3887 (56.5%)  Post-operative and survival data The median length of stay (LOS) was 8 days. In-hospital mortality was 2.9%. Among patients who survived the index hospitalization, most were discharged home (70.8%), while 26.3% were discharged to rehabilitation. Kaplan–Meier estimates of survival were 78.8 and 43.5% at 1 and 5 years, respectively. Survival was better for females than for males (5-year survival: 47.9 vs. 33.6%, P < 0.001) and decreased with age (5-year survival: 63.1, 49.6 and 28.9% for age groups of 65–74, 75–84 and >85 years, respectively, P < 0.001). Healthcare utilization and healthcare-related costs Figure 1 demonstrates the changes in mean monthly healthcare-related costs in the base year and the follow-up year, in USD. Total mean monthly costs increased by 96% ($1 634 vs. $833, 5 719 vs. 2914 NIS, P < 0.001). Most of this increase was attributed to a 2 082% increase in mean monthly costs for rehabilitation ($660 vs. $30, 2 310 vs. 105 NIS, P < 0.001). Excluding rehabilitation costs, mean monthly costs were still 21% higher in the follow-up year ($975 vs. $803, 3 413 vs. 2 811 NIS, P < 0.001). This increase was attributed to an increase in inpatient costs (11% increase, $750 vs. $678, 2 624 vs. 2 373 NIS, P < 0.001), outpatient costs (160% increase, $134 vs. $52, 468 vs. 180 NIS, P < 0.001), day-care admissions (11% increase, $59 vs. $53, 205 vs. 184 NIS, P < 0.001), ED visits (27% increase, $12 vs. $9.4, 42 vs 33 NIS, P < 0.001), costs for laboratories (27% increase, $6.9 vs. $5.4, 24 vs. 19 NIS, P < 0.001) and medications costs (50% increase, $1.7 vs. $1.1, 6 vs. 4 NIS, P < 0.001, Fig. 1). Figure 1 View largeDownload slide Mean monthly healthcare-related costs (in 2017 USD) before and after hip fracture, according to healthcare sub-categories (1 USD = 3.5 New Israeli Shekels). Figure 1 View largeDownload slide Mean monthly healthcare-related costs (in 2017 USD) before and after hip fracture, according to healthcare sub-categories (1 USD = 3.5 New Israeli Shekels). Healthcare expenditure was higher for males; mean monthly costs in the base year were $960 (3361 NIS) and $776 (2715 NIS) for males and females, respectively (P < 0.001). During the follow-up year, a 107 and 91% increase in total mean monthly costs was seen for males and females, respectively (males: 1987$ [6955 NIS], females: 1482$ [5185 NIS], P < 0.001). The highest increase in costs (138%) was seen in patients aged over 85 years ($1 730 [6056 NIS] vs. $726 [2541 NIS], P < 0.001). Costs increased by 82 and 55% in the 75–84 age group (1562$ [5 469 NIS] vs. 858$ [3002 NIS]) and 65–74 age group (1 593$ [5576 NIS] vs. 1027$ [3593 NIS]), respectively. Healthcare expenditure was higher for patients who died during the follow-up year (mean monthly cost of $4861 [17 013 NIS], 380% increase from the baseline expenditure of 1 012$ [3541 NIS]). In patients who survived the follow-up year, total mean monthly costs increased by 13% ($890 [3.116 NIS] vs. 791$ [2769 NIS]). Factors associated with costs The univariate analysis for mean monthly costs in the follow-up year is presented in Table 2 Factors consistently associated with increased costs were male sex, higher CCI, liver cirrhosis, diabetes, hypertension, current smoking, history of stroke, medications (corticosteroids, progesterone, proton-pump inhibitors and vitamin K antagonists), baseline costs in the year before admission, the location of the fracture, length of surgery, time to surgery, LOSs in the orthopedics and intensive care unit, re-operation, costs of the index hospitalization and mortality (Table 2). Table 2 Univariate associations with healthcare-related costs during the follow-up year, in 2017 USD (1 USD = 3.5 New Israeli Shekels) Variable  Q1 (25th percentile)  Median (50th percentile)  Q3 (75th percentile)  Coefficient (95% CI)  P  Coefficient (95% CI)  P  Coefficient (95% CI)  P  Age (per year)  0.3 (−1.5, 2.0)  0.72  −1.2 (−3.9, 1.5)  0.38  1.8 (−8.9, 12.4)  0.75  Male  57.2 (29.1, 85.2)  <0.001  177.9 (132.8, 223.1)  <0.001  704.0 (544.4, 863.6)  <0.001  Charlson’s comorbidity index score (per point)  37.1 (28.4, 38.3)  <0.001  77.1 (68.7, 85.6)  <0.001  260.4 (238.3, 282.5)  <0.001  Cirrhosis  254.5 (141.3, 367.7)  <0.001  693.7 (509.1, 878.3)  <0.001  2 290 (1 573, 3 006)  <0.001  Diabetes  49.6 (21.5, 77.7)  0.001  122.0 (78.1, 166.0)  <0.001  558.5 (416.1, 701.0)  <0.001  Dementia  −128.8 (−164.9, −92.5)  <0.001  −126.1 (−173.6, −78.6)  <0.001  −125.2 (−302.6, 52.0)  0.17  Parkinson’s disease  39.9 (−8.4, 88.2)  0.11  107.9 (29.8, 186.1)  0.007  162.9 (−141.4, 467.4)  0.29  Hypertension  73.5 (38.7, 108.3)  <0.001  133.6 (84.5, 182.7)  <0.001  342.0 (146.8, 537.3)  0.001  Smoking  57.1 (26.1, 88.2)  <0.001  200.3 (153.3, 247.4)  <0.001  567.5 (407.4, 727.6)  <0.001  Previous stroke  67.7 (37.0, 98.3)  <0.001  158.4 (111.0, 205.9)  <0.001  500.3 (324.8, 675.8)  <0.001  Glucocorticosteroids  74.8 (43.6, 105.9)  <0.001  172.4 (122.7, 222.1)  <0.001  408.1 (237.9, 578.4)  <0.001  Heparin  49.1 (−0.4, 98.6)  0.052  221.3 (144.0, 298.5)  <0.001  854.9 (575.3, 1 134.5)  <0.001  Progesterone  233.0 (31.3, 434.7)  0.024  490.6 (172.6, 808.6)  0.002  2 079 (797.4, 3 360)  0.001  Proton-pump inhibitors  72.7 (46.7, 98.8)  <0.001  139.1 (96.4, 181.8)  <0.001  344.0 (186.9, 501.1)  <0.001  Vitamin K antagonists  135.6 (96.6, 174.8)  <0.001  275.3 (213.2, 337.3)  <0.001  863.6 (636.9, 1 090)  <0.001  Total mean monthly cost—base year (per 100 USD)  14.8 (13.2, 16.5)  <0.001  45.4 (42.9, 47.8)  <0.001  94.5 (87.3, 101.8)  <0.001  Sub-trochanteric fracture  60.5 (13.1, 107.7)  0.012  93.5 (16.1, 170.8)  0.018  112.5 (−206.7, 399.5)  0.53  Direct admission to orthopedics  −170.5 (−213.9, −127.2)  <0.001  −270.6 (−340.9, −200.4)  <0.001  −836.4 (−1 078, −594.3)  <0.001  Length of surgery (per hour)  62.1 (43.1, 81.1)  <0.001  89.0 (54.2, 123.9)  <0.001  192.0 (85.5, 298.6)  <0.001  Underwent surgery  177.4 (136.1, 218.7)  <0.001  −180.8 (−254.1, −107.5)  <0.001  −828.2 (−1 096, −560.5)  <0.001  Surgery within 48 h  36.2 (9.1, 63.3)  0.009  126.6 (82.7, 170.5)  <0.001  509.9 (349.1, 670.8)  <0.001  Time to surgery (per hour)  0.9 (0.6, 1.1)  <0.001  2.0 (1.6, 2.4)  <0.001  7.5 (6.0, 8.9)  <0.001  Total hip replacement  −173.4 (−268.1, −78.8)  <0.001  −240.8 (−383.0, −98.7)  0.001  −616.4 (−1 171, −62.5)  0.029  LOS (per day)  4.4 (2.8, 6.0)  <0.001  20.1 (18.0, 22.1)  <0.001  65.3 (56.3, 74.4)  <0.001  ICU admission  −195.2 (−237.6, −152.8)  <0.001  −10.2 (−71.6, 51.1)  0.74  6.0 (−241.4, 253.3)  0.96  Intensive care unit LOS (per day)  45.7 (30.3, 61.1)  <0.001  270.5 (243.3, 297.7)  <0.001  400.7 (187.1, 614.4)  <0.001  Orthopedics LOS (per day)  13.7 (10.9, 16.5)  <0.001  18.2 (14.2, 22.3)  <0.001  37.8 (22.0, 53.6)  <0.001  Re-operation  264.6 (193.4, 336.5)  <0.001  1 091 (964.0,1217)  <0.001  3304 (2 841, 3 769)  <0.001  Total admission cost (per 100 USD)  3.7 (3.4, 4.1)  <0.001  4.8 (4.3, 5.3)  <0.001  9.4 (7.4, 11.4)  <0.001  Discharge to rehabilitation  219.7 (187.7, 251.7)  <0.001  185.8 (133.6, 238.0)  <0.001  242.1 (66.5, 417.6)  0.007  Mortality at 1 year  1 176 (1 135, 1 216)  <0.001  3094 (3054.4, 3 134)  <0.001  5956 (5860, 6 052)  <0.001  Died during follow-up  130.4 (98.7, 162.1)  <0.001  487.8 (450.9, 524.7)  <0.001  2289 (2200, 2 378)  <0.001  Variable  Q1 (25th percentile)  Median (50th percentile)  Q3 (75th percentile)  Coefficient (95% CI)  P  Coefficient (95% CI)  P  Coefficient (95% CI)  P  Age (per year)  0.3 (−1.5, 2.0)  0.72  −1.2 (−3.9, 1.5)  0.38  1.8 (−8.9, 12.4)  0.75  Male  57.2 (29.1, 85.2)  <0.001  177.9 (132.8, 223.1)  <0.001  704.0 (544.4, 863.6)  <0.001  Charlson’s comorbidity index score (per point)  37.1 (28.4, 38.3)  <0.001  77.1 (68.7, 85.6)  <0.001  260.4 (238.3, 282.5)  <0.001  Cirrhosis  254.5 (141.3, 367.7)  <0.001  693.7 (509.1, 878.3)  <0.001  2 290 (1 573, 3 006)  <0.001  Diabetes  49.6 (21.5, 77.7)  0.001  122.0 (78.1, 166.0)  <0.001  558.5 (416.1, 701.0)  <0.001  Dementia  −128.8 (−164.9, −92.5)  <0.001  −126.1 (−173.6, −78.6)  <0.001  −125.2 (−302.6, 52.0)  0.17  Parkinson’s disease  39.9 (−8.4, 88.2)  0.11  107.9 (29.8, 186.1)  0.007  162.9 (−141.4, 467.4)  0.29  Hypertension  73.5 (38.7, 108.3)  <0.001  133.6 (84.5, 182.7)  <0.001  342.0 (146.8, 537.3)  0.001  Smoking  57.1 (26.1, 88.2)  <0.001  200.3 (153.3, 247.4)  <0.001  567.5 (407.4, 727.6)  <0.001  Previous stroke  67.7 (37.0, 98.3)  <0.001  158.4 (111.0, 205.9)  <0.001  500.3 (324.8, 675.8)  <0.001  Glucocorticosteroids  74.8 (43.6, 105.9)  <0.001  172.4 (122.7, 222.1)  <0.001  408.1 (237.9, 578.4)  <0.001  Heparin  49.1 (−0.4, 98.6)  0.052  221.3 (144.0, 298.5)  <0.001  854.9 (575.3, 1 134.5)  <0.001  Progesterone  233.0 (31.3, 434.7)  0.024  490.6 (172.6, 808.6)  0.002  2 079 (797.4, 3 360)  0.001  Proton-pump inhibitors  72.7 (46.7, 98.8)  <0.001  139.1 (96.4, 181.8)  <0.001  344.0 (186.9, 501.1)  <0.001  Vitamin K antagonists  135.6 (96.6, 174.8)  <0.001  275.3 (213.2, 337.3)  <0.001  863.6 (636.9, 1 090)  <0.001  Total mean monthly cost—base year (per 100 USD)  14.8 (13.2, 16.5)  <0.001  45.4 (42.9, 47.8)  <0.001  94.5 (87.3, 101.8)  <0.001  Sub-trochanteric fracture  60.5 (13.1, 107.7)  0.012  93.5 (16.1, 170.8)  0.018  112.5 (−206.7, 399.5)  0.53  Direct admission to orthopedics  −170.5 (−213.9, −127.2)  <0.001  −270.6 (−340.9, −200.4)  <0.001  −836.4 (−1 078, −594.3)  <0.001  Length of surgery (per hour)  62.1 (43.1, 81.1)  <0.001  89.0 (54.2, 123.9)  <0.001  192.0 (85.5, 298.6)  <0.001  Underwent surgery  177.4 (136.1, 218.7)  <0.001  −180.8 (−254.1, −107.5)  <0.001  −828.2 (−1 096, −560.5)  <0.001  Surgery within 48 h  36.2 (9.1, 63.3)  0.009  126.6 (82.7, 170.5)  <0.001  509.9 (349.1, 670.8)  <0.001  Time to surgery (per hour)  0.9 (0.6, 1.1)  <0.001  2.0 (1.6, 2.4)  <0.001  7.5 (6.0, 8.9)  <0.001  Total hip replacement  −173.4 (−268.1, −78.8)  <0.001  −240.8 (−383.0, −98.7)  0.001  −616.4 (−1 171, −62.5)  0.029  LOS (per day)  4.4 (2.8, 6.0)  <0.001  20.1 (18.0, 22.1)  <0.001  65.3 (56.3, 74.4)  <0.001  ICU admission  −195.2 (−237.6, −152.8)  <0.001  −10.2 (−71.6, 51.1)  0.74  6.0 (−241.4, 253.3)  0.96  Intensive care unit LOS (per day)  45.7 (30.3, 61.1)  <0.001  270.5 (243.3, 297.7)  <0.001  400.7 (187.1, 614.4)  <0.001  Orthopedics LOS (per day)  13.7 (10.9, 16.5)  <0.001  18.2 (14.2, 22.3)  <0.001  37.8 (22.0, 53.6)  <0.001  Re-operation  264.6 (193.4, 336.5)  <0.001  1 091 (964.0,1217)  <0.001  3304 (2 841, 3 769)  <0.001  Total admission cost (per 100 USD)  3.7 (3.4, 4.1)  <0.001  4.8 (4.3, 5.3)  <0.001  9.4 (7.4, 11.4)  <0.001  Discharge to rehabilitation  219.7 (187.7, 251.7)  <0.001  185.8 (133.6, 238.0)  <0.001  242.1 (66.5, 417.6)  0.007  Mortality at 1 year  1 176 (1 135, 1 216)  <0.001  3094 (3054.4, 3 134)  <0.001  5956 (5860, 6 052)  <0.001  Died during follow-up  130.4 (98.7, 162.1)  <0.001  487.8 (450.9, 524.7)  <0.001  2289 (2200, 2 378)  <0.001  ICU, intensive care unit; LOS, length of stay; USD, United States dollars. A multivariate quantile regression model for mean monthly costs in the follow-up year is presented in Table 3. Factors consistently associated with increased mean monthly costs were CCI, hypertension, baseline expenditure in the base year, the location of the fracture, the hospital LOS, discharge to a rehabilitation facility and mortality (Table 3). Table 3 Multivariate quantile regression model for mean monthly costs in the follow-up year, in 2017 USD (1 USD = 3.5 New Israeli Shekels) Variable  Q1 (25th percentile)  Median (50th percentile)  Q3 (75th percentile)  Coefficient (95% CI)  P  Coefficient (95% CI)  P  Coefficient (95% CI)  P  Age (per year)  −2.1 (−4.0, −0.2)  0.032  −1.9 (−4.3, 0.6)  0.134  −5.7 (−10.1, −1.2)  0.012  Male  8.2 (−20.1, 36.6)  0.570  38.1 (−0.1, 76.3)  0.051  108.9 (40.4, 177.4)  0.002  Charlson comorbidity index score (per point)  22.1 (16.7, 27.4)  <0.001  23.5 (16.1, 30.7)  <0.001  55.1 (41.7, 68.7)  <0.001  Cirrhosis  204.9 (82.9, 327.1)  0.001  247.1 (81.6, 412.5)  0.003  –    Dementia  −161.9 (−193.8, −130.0)  <0.001  −196.9 (−240.7, −153.0)  <0.001  −207.1 (−286.1, −128.0)  <0.001  Hypertension  65.7 (32.9, 98.5)  <0.001  63.5 (19.2, 107.7)  0.005  106.5 (26.3, 186.5)  0.009  Parkinson’s disease  –    95.5 (26.9, 163.9)  0.006  –    Vitamin K antagonists  45.3 (5.4, 85.2)  0.026  74.8 (20.8, 128.8)  0.007  –    Total mean monthly cost—base year (per 100 SD)  2.9 (2.4, 3.4)  <0.001  26.7 (24.7, 28.8)  <0.001  53.7 (49.8, 57.6)  <0.001  Sub-trochanteric fracture  64.7 (15.6, 113.8)  0.010  113.6 (47.1, 179.9)  0.001  154.4 (34.3, 274.4)  0.012  Underwent surgery  −406.3 (−692.0, −120.7)  0.005  –    –    Total hip replacement  −101.8 (−197.2, −6.4)  0.036  −226.1 (−355.7, −96.5)  0.001  −369.4 (−593.5, −123.1)  0.003  Direct admission to Orthopedics  −359.1 (−420.2, −297.9)  <0.001  −691.3 (−765.7, −616.8)  <0.001  −1 131.9 (−1 256.1, −1 007.6)  <0.001  LOS (per day)  3.0 (1.2, 4.7)  0.001  15.6 (13.2, 17.9)  <0.001  33.3 (29.0, 37.7)  <0.001  ICU admission  −246.0 (−304.8, −187.0)  <0.001  −490.1 (−564.0, −416.3)  <0.001  −754.0 (−881.7, −626.3)  <0.001  Discharge to rehabilitation  212.7 (183.9, 241.6)  <0.001  143.4 (104.4, 182.5)  <0.001  191.6 (120.8, 262.5)  <0.001  1-year mortality  1 203.7 (1 169, 1 239)  <0.001  3 044.2 (2 996, 3 091)  <0.001  5 531.6 (5 448, 5 615)  <0.001  Constant  1 002    950.2    1 632    Variable  Q1 (25th percentile)  Median (50th percentile)  Q3 (75th percentile)  Coefficient (95% CI)  P  Coefficient (95% CI)  P  Coefficient (95% CI)  P  Age (per year)  −2.1 (−4.0, −0.2)  0.032  −1.9 (−4.3, 0.6)  0.134  −5.7 (−10.1, −1.2)  0.012  Male  8.2 (−20.1, 36.6)  0.570  38.1 (−0.1, 76.3)  0.051  108.9 (40.4, 177.4)  0.002  Charlson comorbidity index score (per point)  22.1 (16.7, 27.4)  <0.001  23.5 (16.1, 30.7)  <0.001  55.1 (41.7, 68.7)  <0.001  Cirrhosis  204.9 (82.9, 327.1)  0.001  247.1 (81.6, 412.5)  0.003  –    Dementia  −161.9 (−193.8, −130.0)  <0.001  −196.9 (−240.7, −153.0)  <0.001  −207.1 (−286.1, −128.0)  <0.001  Hypertension  65.7 (32.9, 98.5)  <0.001  63.5 (19.2, 107.7)  0.005  106.5 (26.3, 186.5)  0.009  Parkinson’s disease  –    95.5 (26.9, 163.9)  0.006  –    Vitamin K antagonists  45.3 (5.4, 85.2)  0.026  74.8 (20.8, 128.8)  0.007  –    Total mean monthly cost—base year (per 100 SD)  2.9 (2.4, 3.4)  <0.001  26.7 (24.7, 28.8)  <0.001  53.7 (49.8, 57.6)  <0.001  Sub-trochanteric fracture  64.7 (15.6, 113.8)  0.010  113.6 (47.1, 179.9)  0.001  154.4 (34.3, 274.4)  0.012  Underwent surgery  −406.3 (−692.0, −120.7)  0.005  –    –    Total hip replacement  −101.8 (−197.2, −6.4)  0.036  −226.1 (−355.7, −96.5)  0.001  −369.4 (−593.5, −123.1)  0.003  Direct admission to Orthopedics  −359.1 (−420.2, −297.9)  <0.001  −691.3 (−765.7, −616.8)  <0.001  −1 131.9 (−1 256.1, −1 007.6)  <0.001  LOS (per day)  3.0 (1.2, 4.7)  0.001  15.6 (13.2, 17.9)  <0.001  33.3 (29.0, 37.7)  <0.001  ICU admission  −246.0 (−304.8, −187.0)  <0.001  −490.1 (−564.0, −416.3)  <0.001  −754.0 (−881.7, −626.3)  <0.001  Discharge to rehabilitation  212.7 (183.9, 241.6)  <0.001  143.4 (104.4, 182.5)  <0.001  191.6 (120.8, 262.5)  <0.001  1-year mortality  1 203.7 (1 169, 1 239)  <0.001  3 044.2 (2 996, 3 091)  <0.001  5 531.6 (5 448, 5 615)  <0.001  Constant  1 002    950.2    1 632    ICU, intensive care unit; LOS, length of stay; USD, United States dollars. Discussion Our results demonstrate that hip fracture in the elderly population in Israel is associated with significantly increased healthcare utilization and costs. During the year following a hip fracture, mean monthly costs increased by 96% in comparison to the year before the fracture. Rehabilitation costs comprised of ~40% of the total mean monthly costs during the follow-up year. The increments in healthcare expenditures were higher in males as well as in patients who died during the follow-up year, and increased with age. Multivariate regression identified factors associated with increased mean monthly costs during the follow-up year, which included higher CCI score, certain comorbidities, expenditure in the base year, discharge to a rehabilitation facility and mortality. These data correspond with previous studies [2, 5, 7, 8, 15]. A recent study which examined 6019 patients with hip fracture and assessed costs before and after hospital admission [11], showed that healthcare expenditure increased by 147% during the year following the fracture. Of these costs, 64% were attributed to the acute care of the fracture and 36% to rehabilitation [11]. Similarly, a Canadian study that compared mean annual costs between patients with hip fracture and a matched cohort showed that costs in the hip fracture group increased by 241 and 266% in females and males, respectively [12]. The largest component of costs attributable to the hip fracture was acute hospitalizations, accounting for ~40% (38–41%). Similar to our study, higher costs were demonstrated in patients who died during follow-up. Another recent study analyzed healthcare utilization and costs in 8028 patients with hip fracture in the UK [16]. Mean annual costs increased by ~135%. Inpatient costs comprised 92% of the total costs during the follow-up year (vs. 83% in the base year). This study also demonstrated that variables associated with significantly higher healthcare costs in the follow-up period after hospitalization included age >65 years, higher CCI score, discharge to another institution, hospitalizations and outpatients visits in the baseline year and male sex [16]. Another American study [17] demonstrated a 400% increase in mean healthcare expenditures in the year following hip fracture. Approximately 88% of this increment has been directly related to the hip fracture. Healthcare expenditures increased in all sub-categories of healthcare utilization, except cancer treatment-related costs. These include costs for treatment directly related to the fracture (aftercare, complications, decubitus ulcers) and costs for treatment unrelated to the fracture (cardiovascular, respiratory) [17]. Although differences exist in study designs and setting, all these studies show similar trends: healthcare utilization is at least doubled in the year following a hip fracture. Indeed, in the present study, costs for rehabilitation accounted for <50% of the total costs during the follow-up year. Some of the economic effects demonstrated in this study are quite expected. Hip fracture requires rehabilitation in the majority of cases. Functional health status limitations were previously [5] found in all ADL categories among patients with hip fracture in comparison to controls. Thus, it is no surprise that costs for rehabilitation increased dramatically in the present study. However, we have shown that mean monthly costs have increased in all sub-categories of healthcare utilization during the follow-up year, with a 21% increase in non-rehabilitation costs. A question arises regarding how much of this increase in costs is attributed to the fracture itself. Following the hip fracture, 51% of the patients were found to be physically frail and 34% were disabled (defined as any ADL dependency) [18]. Some degree of cognitive impairment was found in 89% of the patients. Complications were seen in 71% of patients with frailty and cognitive impairment [18]. It was recently shown that patients with frailty utilize healthcare services to a greater extent in comparison to matched patients without frailty [19]. This includes increased general practitioner and outpatient visits, hospitalizations and nursing care. Assuming that a certain proportion of patients have physical and/or cognitive frailty following a hip fracture, it is possible that frailty accounts for a portion of the increased costs during the following year. Therefore, frailty should be a target for rehabilitation, monitoring and treatment, both from the patient safety and quality of life perspective, and from the economic perspective. Another factor to be considered is aging. According to Clalit HMO’s database, healthcare utilization in the general population increases in ~2% per year, due to aging alone. Therefore, the much higher increase in costs seen in the present study is unlikely to be explained by aging alone. Patients who died during the follow-up year had the largest increase in healthcare-related costs. One can assume that they were in a poorer post- and maybe pre-fracture medical condition, and probably required more intensive medical care following the fracture. The present study has some strengths, including a large sample size, accurate billing information collected directly from an administrative database, complete follow-up for all Clalit enrollees included in the study, both in the hospital and in the community settings, and a unique opportunity to explore this topic in a public healthcare system with government dictated healthcare services coverage. Our study has several limitations. Costs were calculated according to Clalit’s database, which includes only direct costs. These costs are either calculated based on the LOS with a per-diem cost, or based on a prospective payment diagnosis-related group-like mechanism, where several procedure codes are compensated on a global basis. However, this method does not take into account detailed costs of procedures (e.g. magnetic resonance imaging) or costly medications (e.g. broad-spectrum antibiotics) given during hospital admission. In addition, indirect costs, such as loss of work days, home assistance that may be required for some patients, and supportive medical devices not covered by the HMO, were not taken into account in our calculations. No quality of life data were obtained, thus the impact of hip fracture on this parameter and interrelations between it and healthcare utilization were not assessed. Some Clalit hospitals have a rehabilitation facility within them. Thus, in some hospitals, costs for rehabilitation (at least partially) were included in the costs for the index hospitalization. This means that at least some of the costs for the index hospitalization are actually attributed to rehabilitation. The study included only Clalit enrollees; however, since most of the basic medical coverage in Israel is dictated by the Ministry of Health, we can assume that costs for these services apply to the other three HMOs in Israel, as well. Finally, some demographic differences between Clalit enrollees and enrollees in the other three HMOs cannot be ruled out. To conclude, our study demonstrates that hip fracture in patients over 65 years of age is associated with a significant increase in total mean monthly costs during the follow-up year, for all examined sub-categories of costs. As the population worldwide is aging and life expectancy increases, it is reasonable to assume that the incidence of hip fractures, and the economic burden it brings, will only increase in the future. Additional study is therefore required to further characterize this increase in costs and to find and define modifiable factors that may affect these changes in healthcare utilization following a hip fracture. Acknowledgements This study was conducted as part of the requirements for graduation from the medical school of the Faculty of Health Sciences, Ben-Gurion University of the Negev. The study was presented at the 33rd conference of the International Society for Quality in Healthcare, Tokyo 2016. References 1 Nayagam S. Injuries of the hip and femur. In: Solomon L, Waraick D, Nayagam S (eds). Apley’s System of Orthopaedics and Fractures , 9th edn. Bristol, UK: Hodder Arnold, 2010: 843– 74. Google Scholar CrossRef Search ADS   2 Brauer CA, Coca-Perraillon M, Cutler DM et al.  . Incidence and mortality of hip fractures in the United States. J Am Med Assoc  2009; 302: 1573– 9. Google Scholar CrossRef Search ADS   3 Carinci F, Van Gool K, Mainz J et al.  . Towards actionable international comparisons of health system performance: expert revision of the OECD framework and quality indicators. Int J Qual Health Care  2015; 27: 137– 46. Google Scholar PubMed  4 Haentjens P, Magaziner J, Colon-Emeric CS et al.  . Meta-analysis: excess mortality after hip fracture among older women and men. Ann Intern Med  2010; 152: 380– 90. Google Scholar CrossRef Search ADS PubMed  5 Wolinsky FD, Fitzgerald JF, Stump TE. The effect of hip fracture on mortality, hospitalization, and functional status: a prospective study. Am J Public Health  1997; 87: 398– 403. Google Scholar CrossRef Search ADS PubMed  6 LeBlanc ES, Hillier TA, Pedula KL et al.  . Hip fracture and increased short-term but not long-term mortality in healthy older women. Arch Intern Med  2011; 170: 1831– 7. Google Scholar CrossRef Search ADS   7 Tosteson ANA, Gottlieb DJ, Radley D et al.  . Excess mortality following hip fracture: the role of underlying health status. Osteoporosis Int  2007; 18: 1463– 72. Google Scholar CrossRef Search ADS   8 Bentler SE, Liu L, Obrizan M et al.  . The aftermath of hip fracture: discharge placement, functional status change, and mortality. Am J Epidemiol  2009; 170: 1290– 9. Google Scholar CrossRef Search ADS PubMed  9 Johnell O, Kanis JA. An estimate of the worldwide prevalence, mortality and disability associated with hip fracture. Osteoporosis Int  2004; 15: 897– 902. Google Scholar CrossRef Search ADS   10 Orsini LS, Rousculp MD, Long SR et al.  . Health care utilization and expenditures in the United States: a study of osteoporosis-related fractures. Osteoporosis Int  2005; 16: 359– 71. Google Scholar CrossRef Search ADS   11 Duclos A, Souray-Targe S, Randrianasolo M et al.  . Burden of hip fracture on inpatient care: a before and after population-based study. Osteoporosis Int  2010; 21: 1493– 1501. Google Scholar CrossRef Search ADS   12 Nikitovic M, Wodchis WP, Krahn MD et al.  . Direct health-care costs attributed to hip fractures among seniors: a matched cohort study. Osteoporosis Int  2013; 24: 659– 69. Google Scholar CrossRef Search ADS   13 Grebregziagher M, Lynch CP, Mueller M et al.  . Using quantile regression to investigate racial disparities in medication non-adherence. BMC Med Res Methodol  2011; 11: 88. Google Scholar CrossRef Search ADS PubMed  14 Rehkopf DH. Quantile regression for hypothesis testing and hypothesis screening at the dawn of big data. Epidemiology  2012; 23: 665– 7. Google Scholar CrossRef Search ADS PubMed  15 Panula J, Pihlajamaki H, Mattila VM et al.  . Mortality and cause of death in hip fracture patients aged 65 or older—a population-based study. BMC Musculoskelet Disord  2011; 12: 105. Google Scholar CrossRef Search ADS PubMed  16 Lambrelli D, Burge R, Raluy-Callado M et al.  . Retrospective database study to assess the economic impact of hip fracture in the United Kingdom. J Med Econ  2014; 17: 817– 25. Google Scholar CrossRef Search ADS PubMed  17 Kilgore ML, Curtis JR, Delzell E et al.  . A close examination of healthcare expenditures related to fractures. J Bone Mineral Res  2013; 28: 816– 20. Google Scholar CrossRef Search ADS   18 Kistler EA, Nicholas JA, Kates SL et al.  . Frailty and short-term outcomes in patients with hip fracture. Ger Ortho Surg Rehab  2015; 6: 209– 14. Google Scholar CrossRef Search ADS   19 Coelho T, Paul C, Gobbens RJJ et al.  . Frailty as a predictor of short-term adverse outcomes. PeerJ  2015; 3: e1121. Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2017. Published by Oxford University Press in association with the International Society for Quality in Health Care. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

International Journal for Quality in Health CareOxford University Press

Published: Mar 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 12 million articles from more than
10,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Unlimited reading

Read as many articles as you need. Full articles with original layout, charts and figures. Read online, from anywhere.

Stay up to date

Keep up with your field with Personalized Recommendations and Follow Journals to get automatic updates.

Organize your research

It’s easy to organize your research with our built-in tools.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

Monthly Plan

  • Read unlimited articles
  • Personalized recommendations
  • No expiration
  • Print 20 pages per month
  • 20% off on PDF purchases
  • Organize your research
  • Get updates on your journals and topic searches

$49/month

Start Free Trial

14-day Free Trial

Best Deal — 39% off

Annual Plan

  • All the features of the Professional Plan, but for 39% off!
  • Billed annually
  • No expiration
  • For the normal price of 10 articles elsewhere, you get one full year of unlimited access to articles.

$588

$360/year

billed annually
Start Free Trial

14-day Free Trial