Referral Patterns and Risk of Early Revision after Primary Total Joint Arthroplas

描述（由申请人提供）：本申请涉及挑战领域（05）比较有效性研究（CER）和高优先主题（05- eb -105）医疗植入物的比较有效性。髋关节或膝关节翻修全关节置换术（rTJA）是一项复杂、昂贵的手术。在美国，每年约有83,000例手术（2006年估计）。这些翻修手术的频率预计只会随着未来几年预计进行的大量初级全关节置换术（pTJA）手术而增加（预计到2030年每年将超过400万例）。考虑到rTJA手术的结果比pTJA更差，包括更高的并发症风险、更高的死亡率和更频繁的进一步翻修手术，有必要减少对这些手术的需求，同时也要优化这些翻修手术的结果。本研究的目的是：1)评估患者和制度因素对pTJA后早期rTJA需求的影响；2)确定了第一次pTJA后再tja的转诊模式及其预测因素；3)确定转诊方式对rTJA术后并发症的影响。这些目标将通过在1997年1月1日至2006年12月31日期间确定纽约州和加利福尼亚州的居民进行首次pTJA来实现。这些患者将被随访至研究期结束，对同一关节进行rTJA治疗。那些在不同的外科医生和不同的医院做手术的人将被视为“转诊”。患者将通过纽约和加利福尼亚现有的医院出院数据库进行识别。将评估患者和机构（外科医生和医院）因素的影响。为了实现这3个具体目标，这些效果将根据rTJA的时间、转诊的可能性和rTJA后的并发症进行评估。这将使用不同的多变量模型来分析这3个目标。将使用Cox回归模型分析Aim 1的rTJA时间。目标2和目标3将使用多变量重复测量模型进行评估，包括考虑由同一外科医生或同一医院内进行的手术的相关性的方法。目标2的模型将评估转诊的预测因素。目的3的模型将确定转诊是否能改善短期结果，特别是手术并发症、90天再入院率、90天住院死亡率和随后的翻修手术。这些数据将在转诊患者和未转诊患者之间进行比较，并对潜在的其他混杂因素进行调整。从美国人群的角度来看，目前尚不清楚哪些患者早期rTJA的风险最大，特别是考虑到患者和机构因素的结合。早期rTJA的转诊模式也是未知的。在rTJA方面是否存在地理差异也是未知的。最后，目前尚不清楚转介早期rTJA的患者是否比那些在小容量外科医生或中心接受手术的患者有更好的结果。所有这些问题都将在本研究中进行探讨。快速识别早期rTJA风险最大的患者将有助于未来研究的发展，并有希望通过临床干预来预防这些早期失败。rTJA的转诊模式和趋势直接影响到医疗资源的分配，特别是涉及到培训和招募接受过奖学金培训的关节置换外科医生。此外，外科医生或医院对早期rTJA术后结果的影响与提高护理质量和降低与这些复杂且已经昂贵的手术相关的成本直接相关。这个项目解决了我们对rTJA的理解中一个极其重要的问题。所获得的知识将使我们能够进一步减少翻修关节置换术的需要，并通过减少并发症和优化资源分配来确定改善翻修手术后结果的方法。强大的调查团队和一流的学术环境保证了高质量的研究结果。对行政数据的分析是在利用这些数据评价与政策有关的问题方面向前迈出的一步。先前关于骨科结果的研究主要集中在关节置换术后体积结果与并发症或死亡率的关系上，而本研究在检查翻修手术的转诊模式和这些转诊模式的含义方面进入了新的领域。这代表了以前用于这类大型数据库分析的方法的实质性改进，并将为今后的这类分析提供模板。翻修性关节置换术通常是一项复杂、昂贵的手术，其结果比初次关节置换术差。翻修手术被认为更常在三级护理中心和大容量外科医生进行，但转诊模式实际上是未知的。本研究旨在确定1996年至2007年间纽约和加利福尼亚转介翻修关节置换术的模式，并确定早期翻修（首次手术10年内）的危险因素和翻修手术后并发症的危险因素。

DESCRIPTION (provided by applicant): This application addresses Challenge Area (05) Comparative Effectiveness Research (CER) and High Priority Topic (05-EB-105) Comparative Effectiveness of Medical Implants. Revision total joint arthroplasty (rTJA) for the hip or knee joint is a complex, costly procedure. Approximately 83,000 are performed annually in the US (2006 estimate). The frequency of these revision procedures is only expected to increase with the large number of primary total joint arthroplasty (pTJA) procedures that are expected to be performed in the coming years (estimated at over 4 million annually by 2030). Given that rTJA procedures have worse outcomes than pTJA including higher risk of complication, higher mortality, and more frequent need for further revision surgery, there is a need to both decrease the need for these procedures, but also to optimize outcomes for these revisions. The aims of this research are to: 1) evaluate the effect of patient, and institutional factors on the need for early rTJA after pTJA; 2) identify the referral patterns for rTJA following first pTJA and predictors of these patterns; and 3) identify the effect of referral patterns on complications after rTJA. These aims will be accomplished by identifying state residents in New York and California undergoing their first pTJA between January 1, 1997 and December 31, 2006. These patients will be followed until the end of the study period for rTJA on the same joint. Those who have surgery with a different surgeon and at a different hospital will be considered "referrals". Patients will be identified through existing hospital discharge databases for New York and California. The effects of patient and institutional (surgeon and hospital) factors will be evaluated. In order to address the 3 specific aims, these effects will be evaluated for time to rTJA, likelihood of referral, and complications following rTJA. This will be analyzed using various multivariable models for each of the 3 aims. A Cox Regression model will be used to analyze time to rTJA for Aim 1. Aims 2 and 3 will be evaluated using a multivariable repeated measures models including methods that account for the correlation of procedures performed by the same surgeon or within the same hospital. The model for Aim 2 will evaluate the predictors of referral. The model for Aim 3 will determine whether referral improves short-term outcomes, particularly rates of surgical complication, 90 day readmission, 90 day in-hospital mortality, and subsequent revision surgery. These will be compared between patients who are referred and those who are not, adjusting for potential other confounders. It is currently unknown which patients are most at risk of early rTJA from a population perspective in the US, particularly with regard to the combination of patient and institutional factors. It is also unknown what the referral patterns are for early rTJA. It is also unknown whether geographic variation exists with regard to rTJA. Finally, it is unknown whether patients who are referred for early rTJA have better outcomes than those who have their surgery with lower volume surgeons or centers. All of these questions will be explored in this study. Quickly identifying patients most at risk for early rTJA will assist in development of future research and hopefully clinical interventions to prevent these early failures. Referral patterns and trends for rTJA have direct implications for allocation of medical resources, particularly as it pertains to training and recruitment of fellowship trained arthroplasty surgeons. Furthermore, the effect of surgeon or hospital on outcomes after early rTJA are directly relevant to improving the quality of care and reducing costs associated with these complex and already costly procedures. This project addresses an extremely important problem in our understanding of rTJA. The knowledge gained will allow us to move forward to reduce need for revision arthroplasty and to identify ways to improve outcomes following revision surgery by reducing complications and optimizing resource allocation. The strong investigative team and premier academic environment in which this study will take place assures that a high quality research will result. This analysis of administrative data represents a step forward in the use of these data to evaluate policy related questions. While previous studies in orthopedic outcomes have focused on volume outcome relationships and complications or mortality following joint replacement, this analysis moves into new territory in examining the referral patterns for revision surgery and the implications of these patterns of referral. This represents a substantial refinement in the methodology previously used for these kinds of large database analyses and will provide a template for future analyses of this kind. Revision arthroplasty is often a complex, costly procedure with worse outcomes than primary arthroplasty. Revision procedures are thought to be more commonly performed at tertiary care centers and by higher volume surgeons, but referral patterns are actually unknown. This study proposes to identify patterns for referral for revision arthroplasty in New York and California between 1996 and 2007 and to identify risk factors for early revision (within 10 years of primary surgery) and risk factors for complication following revision surgery.