|Home | About | Journals | Submit | Contact Us | Français|
The purpose of this study was to explore the increasing prevalence of factors affecting hospital charges for primary total hip replacement/total knee replacement (THR/TKR). This study analysed 37,918 THR and 76,727 TKR procedures performed in Taiwan from 1996 to 2004. Odds ratio (OR) and effect size (ES) were calculated to assess the relative change rate. Multiple regression models were employed to predict hospital charges. The following factors were associated with increased hospital charges: age younger than 65 years old; increased disease severity (Charlson comorbidity index [CCI]=1 or ≥2); absence of primary diagnoses of osteoarthritis (OA), rheumatoid arthritis (RA), avascular necrosis (AVN); treatment at a hospital or by a surgeon performing a high volume of operations; and longer average length of stay (ALOS). The Bureau of National Health Insurance (BNHI) should ensure that surgeons take precautionary measures to minimise complications and maximise quality of life after surgery. Use of joint prostheses from different manufacturers can reduce costs without compromising patient satisfaction.
Le propos de cette étude est d’explorer les facteurs influençant la durée moyenne de séjour des prothèses totales de hanche et des prothèses totales du genou. Cette étude a analysé 37 918 prothèses totales de hanches et 76 727 prothèses totales du genou, réalisées à Taiwan de 1996 à 2004. Les facteurs suivants étaient associés à une augmentation de la durée de séjour: un âge inférieur à 65 ans. L’indice de sévérité des lésions CCI=1 ou ≥2, l’absence de diagnostic primaire (coxarthrose, arthrite rhumatoïde, nécrose) le traitement dans un hôpital ou par un praticien réalisant de gros volumes opératoires, la longueur de l’ALOS. Le Bureau Central d’Assurance BNHI demande que les praticiens prennent les mesures nécessaires pour diminuer le nombre de complications et augmenter la qualité de vie après l’intervention. L’utilisation de prothèses de différents fabricants permet de réduire les coûts sans compromettre la satisfaction des patients.
The purpose of this study was to explore the increasing prevalence of factors affecting hospital charges of primary total hip replacement/total knee replacement (THR/TKR). This study analysed 37,918 THR and 76,727 TKR procedures performed in Taiwan from 1996 to 2004. Odds ratio (OR) and effect size (ES) were calculated to assess the relative change rate. Multiple regression models were employed to predict hospital charges. The following factors were associated with increased hospital charges: age younger than 65 years old; increased disease severity (Charlson comorbidity index [CCI]=1 or ≥2); absence of primary diagnoses of osteoarthritis (OA), rheumatoid arthritis (RA), avascular necrosis (AVN); treatment at a hospital or by a surgeon performing a high volume of operations; and longer ALOS. The Bureau of National Health Insurance (BNHI) should ensure that surgeons take precautionary measures to minimise complications and maximise quality of life after surgery. Use of joint prostheses from different manufacturers can reduce costs without compromising patient satisfaction.
The effectiveness of total hip replacement/total knee replacement (THR/TKR) surgery in relieving pain and improving physical function has been well documented [1, 2]. In addition to improved surgical techniques, the excellent performance of new implant materials and designs has substantially increased the demand for THR/TKR. The growing population of elderly patients is yet another factor. In November 1997, the Bureau of National Health Insurance (BNHI) in Taiwan implemented a case payment system for THR/TKR in order to control the use of medical resources. Since the system provides insurance coverage for expensive and frequently used medical items, the financial burden of THR/TKR should not be overlooked.
Although the volume of THR/TKR procedures is increasing annually, the prevalence and hospital charges for these procedures have not been documented in a Taiwan study. The BNHI reported 12,088 primary TKR procedures performed in 2004, a 5.8% increase from the 11,425 procedures in 2003; during the same period, 6,519 primary THR procedures were performed which was a 2.1% rate of increase . Thus, this study explored the increasing prevalence of factors affecting hospital charges for THR/TKR.
The source of data analysed was the administrative claims data from the BNHI. Because the BNHI is the sole purchaser in Taiwan, the BNHI dataset was presumed to be the most comprehensive and reliable data source for the study. This study analysed data for all patients who received primary THR/TKR procedures (ICD-9-CM procedure codes of 81.51/81.54) during 1996–2004. All THR/TKR procedures performed for treatment of chronic or complicated diseases and traffic accidents were excluded from analysis. The study analysed 39,569 THR and 76,727 TKR procedures.
This study analysed four dimensions of THR/TKR procedures performed in Taiwan: (1) patient demographic characteristics, (2) patient clinical characteristics, (3) hospital/physician surgical volume, and (4) medical resource usage. Demographic characteristics studied included age and gender. Age categories were 44 years old and younger, 45–64 years old, and 65 years old and older. Clinical characteristics analysed were disease severity and primary diagnoses. Disease severity was classified by Charlson comorbidity index (CCI) scores of 0, 1, or ≥2 . Primary diagnoses were categorised as osteoarthritis (OA), rheumatoid arthritis (RA), avascular necrosis (AVN), and other. The ICD-9-CM codes for primary diagnoses of OA, RA, and AVN were 715.00–715.98; 714 and 714.0; and 733.40, 733.42, and 733.49, respectively. Surgical volume of hospital and surgeon were classified as low (≤25th percentile), medium (26th–74th percentiles), and high (≥75th percentile). Medical resource usage was measured by average length of stay (ALOS).
The study period was divided into three approximately equal time intervals (period 1: 1996–1998; period 2: 1999–2001; and period 3: 2002–2004). Odds ratio (OR) and effect size (ES) were determined to assess the relative change for each factor when using period 1 as the reference group in comparison with period 3. Effect size was calculated as the difference between the mean preoperative score and postoperative score, divided by the standard deviation of the preoperative score . Effect size is a method of standardising the extent of change between different factors. An ES of 1.0 is equivalent to a change of one standard deviation in the sample. Effect sizes of 0.2, 0.5, and 0.8 are typically regarded as indicating small, medium, and large changes, respectively. Multiple regression models were used to predict hospital charges, including the patient demographic and clinical characteristics, hospital/surgeon surgical volume, and medical resource usage. Hospital charges of different levels of hospitals were adjusted for differences in reimbursement from the BNHI, which provides the largest reimbursement to medical centres and the smallest to district hospitals. Statistical analyses were conducted using SPSS version 15.0 (SPSS Inc., Chicago, IL, USA). To reflect the real dollar value, all dollar values at the end of each year were also adjusted to 2004 Taiwan currency values. All hospital charges were then converted from Taiwan dollars to U.S. dollars using an exchange rate of 31.5:1, based on the average exchange rate during 1996–2004. All tests were two-sided, and p values less than 0.05 were considered statistically significant.
The prevalence rate of THR/TKR was 16.69/22.86 per 100,000 persons in 1996 and gradually increased to 22.65/54.95 in 2004, which represented increases at rates of 35.71% and 140.38% separately (Fig. 1). In 1996, the prevalence rate of TKR was 1.72 times that of THR. In 2004, it strikingly increased to 2.43.
Table 1 shows the increasing volume of THR/TKR procedures and the changes in patient demographic and clinical characteristics. Approximately 40% of all THR patients treated from periods 1 to 3 were female, and the number of female patients significantly increased between periods 1 and 3 (OR=1.07, 95% CI=1.02, 1.12). Conversely, the number of male patients significantly decreased (OR=0.95, 95% CI=0.92, 0.99). Approximately 65% of all THR patients were under 64 years of age. The number of THR patients younger than 44 years old significantly decreased between periods 1 and 3 (OR=0.87, 95% CI =0.82, 0.91), but the number of THR patients older than 65 years of age significantly increased (OR=1.17, 95% CI=1.11, 1.23). In terms of clinical characteristics, the data revealed a statistically significant increase in the number of patients with CCI scores of 1 and more between periods 1 and 3 (OR=1.24, 95% CI=1.14, 1.35, and OR=1.31, 95% CI=1.13, 1.51, respectively). The principal diagnoses during periods 1–3 were OA and AVN, which ranged from 39.52–42.52% and from 43.04–47.38%, respectively.
Analysis of demographic characteristics revealed that 70% of all TKR patients were female; the percentage of females increased 1.57% between period 1 and period 3, whereas the percentage of males decreased (−4.25%). Approximately 70% of TKR patients were older than 65 years. The number of patients younger than 44 years old and the number of patients 45–64 years old significantly declined between periods 1 and 3 (OR=0.61, 95% CI=0.52, 0.72 and OR=0.82, 95% CI =0.79, 0.85, respectively), but the number of patients older than 65 years significantly increased (OR=1.09, 95% CI=1.06, 1.12). In clinical terms, the number of patients who had CCI scores of 1 and more significantly increased between periods 1 and 3 (OR=1.25, 95% CI =1.18, 1.32, and OR=1.14, 95% CI=1.02, 1.28, respectively), and the number of patients with CCI scores of 0 also significantly increased (OR=0.96, 95% CI=0.94, 0.99). The principal diagnosis in the TKR patients was OA, which ranged from 91.37% to 94.54%.
Table 2 presents the changing trend in medical resources allocated to THR/TKR. Average length of stay (ALOS) and hospital charges of THR/TKR significantly declined (p<0.05). In THR, ALOS declined from 11.82 days to 8.96 days (−24.24%), and total hospital charges decreased from $4,523.47 to $3,960.06 (−12.46%). In TKR, ALOS decreased from 12.83 days to 9.07 days (−29.31%), and hospital charges decreased from $5,063.17 to $4,330.50 (−14.47%). Additionally, the decline in ES of ALOS and hospital charges was larger in TKR patients than in THR patients between periods 1 and 3.
Tables 3 shows the multiple regression models used to evaluate the predictors for hospital charges. The statistically significant predictors for hospital charges among THR patients were gender, age, severity of disease, primary diagnoses, surgical volumes of surgeon and hospital, and ALOS. This model indicated that male gender, CCI scores of 1 or higher, RA, large surgical volumes of treating surgeon and/or hospital, and longer ALOS were associated with higher hospital charges for THR, while older age was associated with lower hospital charges. Additionally, the statistically significant predictors for increased total hospital charges among TKR patients were age, disease severity, primary diagnoses, surgical volumes of surgeon and hospital, and ALOS (p<0.05). This model indicated that CCI scores of 1 or more, other diagnoses, larger surgical volume of surgeon and hospital, and longer ALOS were associated with higher hospital charges for TKR, while older age was associated with lower hospital charges. All other covariates did not statistically affect hospital charges.
According to the standardised coefficients, ALOS had the largest coefficient with hospital charges while the lowest was a CCI score of 1. Finally, the influence of age on hospital charges in TKR patients cannot be ignored because its standardised coefficient was approximately 0.5 (Table 3).
This population-based study was the first to analyse patient demographics and clinical characteristics, hospital/surgeon surgical volume, and medical resource usage to examine changing trends in the prevalence of THR/TKR surgery and to identify factors predictive of hospital charges of THR/TKR surgery. This study demonstrated that the prevalence of these procedures has increased annually. Likely causes are the growing population of elderly patients and improving surgical techniques, implant materials, and designs [6, 7]. This study confirmed the findings of previous studies [8, 9] that female gender, advanced age (≥65 years old) and CCI sores of 1 or higher were associated with increased likelihood of undergoing THR/TKR.
In Taiwan, the principal diagnoses in THR patients are OA (approximately 40%) and AVN (approximately 45%), and approximately 94% of TKR patients are diagnosed with OA. However, a U.S. study by Bozic et al.  noted that 86% of THR and 90% of TKR patients had OA. A possible explanation for these findings is that THR patients in Taiwan are more likely to use antibiotics, tobacco, and alcohol than those in the United States; however, further research is needed.
Analysis of trends in medical resource usage revealed that ALOS and hospital charges dramatically decreased between period 1 and period 3; this decrease may have been due to improvements in surgical technique, implant material, design, and changes in the medical payment system. Previous studies  reveal that minimally invasive surgery can simultaneously reduce complications as well as duration of recovery. The case payment system in Taiwan had not been implemented until period 3 of this study whereas fee-for-service had been implemented in period 1. Many previous studies [12, 13] indicate that case payment dramatically reduces ALOS and hospital charges simultaneously.
In multiple regression analyses, gender was a risk factor in THR, and, on average, males had higher total hospital charges than females, which may have been due to increased disease severity in males than in females before surgery. This study showed that primary diagnoses of other diseases in younger THR patients were mainly ankylosing spondylitis (4.4%), traumatic arthropathy (2.4%), congenital hip deformity (1.5%), congenital hip dislocation (1.6%), and pyogenic arthritis (1.5%). In younger TKR patients, other primary diagnoses were malignant neoplasm of long bones of lower limb (5.5%), traumatic arthropathy (1.1%), congenital factor VIII disorder (1.0%), gouty arthropathy (0.6%), and chronic osteomyelitis (0.5%). Patients aged younger than 44 had a higher proportion of primary diagnoses than any other age group. A possible cause is the higher hospital charges in this group.
Moreover, higher CCI scores were associated with higher hospital charges. This finding is also consistent with previous studies [14, 15]. Primary diagnosis of other medical conditions was a factor in hospital charges. On further analysis, primary diagnoses of other factors in THR mainly involved mechanical complications of the internal orthopaedic device, implant, or graft (11.5%), fracture (9.3%), congenital hip deformity (7.4%), ankylosing spondylitis (6.6%), and traumatic arthropathy (6.3%). In TKR, other conditions were traumatic arthropathy (11.2%), gouty arthropathy (7.5%), malignant neoplasm of long bones of lower limb (7.0%), mechanical complications of the internal orthopaedic device, implant, or graft (5.3%), and chronic osteomyelitis (4.7%). A previous study  demonstrated that patients with primary diagnoses of other factors tend to have a longer ALOS as well as higher hospital charges.
As hospital charges increase, patients requiring THR/TKR tend to seek treatment from surgeons and hospitals with high surgical volume. Previous studies [16, 17] found that many clinical parameters are strongly associated with hospital charges. However, many clinical parameters could not be analysed in this study due to the limited availability of claims data.
The surgical volume of TKR/THR is currently increasing, and the rate of medical resource usage is rising. This study revealed that the following factors were associated with considerably increased hospital charges: young age; increased disease severity; absence of primary diagnoses of OA, RA, and AVN; higher surgical volume of surgeon or hospital; and a longer ALOS. These factors must be carefully managed to minimise medical resource usage. Additionally, the BNHI must develop a programme for minimising complications and improving quality of life after these procedures. Joint prostheses from different manufacturers must be carefully regulated to maximise patient safety and to ensure efficient use of medical resources.
This work was supported by a grant (NSC 96–2314-B-037–050-MY3) from the National Science Council in Taiwan.