Accuracy of Medicare claims data for rheumatologic diagnoses in total hip replacement recipients

Depression readmission risk is elevated in multiple sclerosis compared to other chronic illnesses

OBJECTIVE

Assess readmissions for depression or suicide attempt (SA) after MS admission versus other chronic inflammatory illnesses.

METHODS

This retrospective cohort study identified MS, asthma, rheumatoid arthritis (RA), depression, and SA in the 2013 National Readmissions Database by International Classification of Diseases codes. Index admissions (MS, n = 7698; asthma, n = 93,590; RA, n = 3685) and depression or SA readmission rates were analyzed. Hazard ratios (HRs) estimated 1-year depression/SA readmission hazard, comparing MS to asthma or RA, adjusting for age, sex, psychiatric comorbidity, substance abuse, tobacco use, income, and index hospitalization characteristics.

RESULTS

MS had more baseline depression (24.7%) versus asthma (15.6%) and RA (14.6%). Ninety-day depression readmission rate was higher in MS (0.5%) than asthma (0.3%) and RA (0.03%). Depression readmission HR was higher after MS admission versus asthma (HR = 1.37, 95% confidence interval (CI) = 1.00-1.86, p = 0.0485) and RA (HR = 4.68, 95% CI = 1.60-13.62, p = 0.0047). HR was not different for SA readmission across groups. Depression readmission HR was more than double in MS patients with psychiatric disease or substance abuse versus RA or asthma patients with either comorbidity.

CONCLUSION

Depression readmission risk after MS hospitalization was elevated versus asthma/RA. Substance use and baseline psychiatric comorbidity were more strongly associated with depression readmission in MS patients.

Claims-based algorithms for common chronic conditions were efficiently constructed using machine learning methods

Identification of medical conditions using claims data is generally conducted with algorithms based on subject-matter knowledge. However, these claims-based algorithms (CBAs) are highly dependent on the knowledge level and not necessarily optimized for target conditions. We investigated whether machine learning methods can supplement researchers' knowledge of target conditions in building CBAs. Retrospective cohort study using a claims database combined with annual health check-up results of employees' health insurance programs for fiscal year 2016-17 in Japan (study population for hypertension, N = 631,289; diabetes, N = 152,368; dyslipidemia, N = 614,434). We constructed CBAs with logistic regression, k-nearest neighbor, support vector machine, penalized logistic regression, tree-based model, and neural network for identifying patients with three common chronic conditions: hypertension, diabetes, and dyslipidemia. We then compared their association measures using a completely hold-out test set (25% of the study population). Among the test cohorts of 157,822, 38,092, and 153,608 enrollees for hypertension, diabetes, and dyslipidemia, 25.4%, 8.4%, and 38.7% of them had a diagnosis of the corresponding condition. The areas under the receiver operating characteristic curve (AUCs) of the logistic regression with/without subject-matter knowledge about the target condition were .923/.921 for hypertension, .957/.938 for diabetes, and .739/.747 for dyslipidemia. The logistic lasso, logistic elastic-net, and tree-based methods yielded AUCs comparable to those of the logistic regression with subject-matter knowledge: .923-.931 for hypertension; .958-.966 for diabetes; .747-.773 for dyslipidemia. We found that machine learning methods can attain AUCs comparable to the conventional knowledge-based method in building CBAs.

Epidemiology of Rheumatoid Arthritis in Korea

Rheumatoid arthritis (RA) is a chronic systemic autoimmune disease characterised by symmetrical involvement of the joints, associated extra-articular manifestations and functional disability. In Korea, several epidemiologic studies reporting prevalence and incidence rates of RA have been conducted using large databases such as claims databases, national surveys, prospective cohort databases or electronic health records; according to these data sources, the estimated prevalence ranged from 0.27% to 1.85%. The prevalence of extra-articular manifestations such as interstitial lung disease (ILD) and Sjögren's syndrome (SS) were also reported, but an issue of external validity of the study results persisted. In this review, we detail the epidemiology of Korean RA patients, focusing on the prevalence of RA and the frequency of systemic extra-articular manifestations including ILD and SS reported in previous studies. In addition, we discuss the current methodological issues which are inherent in Korean epidemiologic studies for patients with RA with understanding of the characteristics of each database.

Validation of Self-Reported Rheumatoid Arthritis Using Medicare Claims: A Nationally Representative Longitudinal Study of Older Adults

Objective

To determine the validity of self‐reported physician diagnosis of rheumatoid arthritis (RA) using multiple gold‐standard measures based on Medicare claims in a nationally representative sample of older adults and to verify whether additional questions about taking medication and having seen a physician in the past two years for arthritis can improve the positive predictive value (PPV) and other measures of the validity of self‐reported RA.

Methods

A total of 3768 Medicare‐eligible respondents with and without incident self‐reported RA were identified from the 2004, 2008, and 2012 waves of the United States Health and Retirement Study. Self‐reported RA was validated using the following three claims‐based algorithms: 1) a single International Classification of Diseases, ninth edition, Clinical Modification claim for RA, 2) two or more claims no greater than 2 years apart, and 3) two or more claims with at least one diagnosis by a rheumatologist. Additional self‐report questions of medication use and having seen a doctor for arthritis in the past two years were validated against the same criteria.

Results

A total of 345 respondents self‐reported a physician diagnosis of RA. Across all three RA algorithms, the PPV of self‐report ranged from 0.05 to 0.16., the sensitivity ranged from 0.23 to 0.55., and the κ statistic ranged from 0.07 to 0.15. Additional self‐report data regarding arthritis care improved the PPV and other validity measures of self‐report; however, the values remained low.

Conclusion

Most older adults who self‐report RA do not have a Medicare claims history consistent with that diagnosis. Revisions to current self‐reported RA questions may yield more valid identification of RA in national health surveys.

Accuracy of diagnoses of inflammatory arthritis in administrative hospitalization databases

PURPOSE

Studies using administrative hospitalization data often classify patients as having inflammatory arthritis based on diagnoses recorded at the hospitalization. We examined the agreement of these diagnoses with patients' prior medical histories.

METHODS

We identified Medicare beneficiaries hospitalized in 2011 to 2015 for total hip arthroplasty (THA), total knee arthroplasty (TKA), acute myocardial infarction (AMI), or sepsis. We compared diagnoses of rheumatoid arthritis (RA) or ankylosing spondylitis (AS) at the index hospitalization to diagnoses over prior inpatient and outpatient claims. To assess the impact of potential misclassification, we compared hospital outcomes using the alternative methods of detecting beneficiaries with arthritis. Analyses were repeated using Medicaid data.

RESULTS

Among 506 781 Medicare beneficiaries with THA, 18282 had RA and 571 had AS at the arthroplasty hospitalization, while 13 212 had RA and 1519 had AS based on claims history. Diagnoses at the hospitalization were highly specific (0.98-0.99), but sensitivities (0.65 for RA; 0.31 for AS) and positive predictive values (PPV) (0.47 for RA; 0.83 for AS) were lower. For TKA, AMI, and sepsis, specificities were 0.97 to 0.99, sensitivities 0.60 to 0.66 for RA and 0.18 to 0.22 for AS, and PPVs 0.43 to 0.47 for RA and 0.73 to 0.77 for AS. In Medicaid, sensitivities were 0.21 to 0.67 for RA and 0.07 to 0.49 for AS. Frequencies of some hospital outcomes differed when arthritis was classified by the index hospitalization or claims history.

CONCLUSION

Diagnoses of RA and AS in hospitalization databases are highly specific but fail to identify large proportions of patients with these diagnoses.

Identification of obstetric anal sphincter injuries (OASIs) and other lacerations: a national survey of nurse-midwives

INTRODUCTION AND HYPOTHESIS

Clinical quality improvement relies on accurate understanding of current practice. We performed a cross-sectional national survey of certified nurse-midwives (CNMs) assessing classification and identification of obstetric anal sphincter injury (OASI) and other delivery lacerations. We hypothesized laceration diagnoses are frequently inaccurate, and delivery records for obstetric lacerations may be of questionable quality.

METHODS

We emailed 6909 American College of Nurse Midwives members an internet-based survey link. Of respondents, we included clinically active CNMs who perform at least one delivery per month. We evaluated laceration knowledge and application using standard descriptive text and images and asked about processes for recording lacerations in the delivery record.

RESULTS

We received 1070 (15.5%) completed surveys and 832 (77.8%) met inclusion criteria. Over 50% characterized their OASI training and ability to identify OASI as good/excellent. Most (79%) had never attended education review on OASI. The overall accuracy for classification and identification of perineal lacerations ranged from 49 to 99%. Non-perineal lacerations were frequently categorized using the perineal/OASI system. Half of respondents (51%) document their deliveries in an electronic medical record but a quarter (28%) are not personally responsible for approving delivery data. Younger participants without a doctoral degree, with self-assessed good/excellent laceration training, and caring for < 50% publicly insured patients had higher accuracy for laceration identification and diagnosis.

CONCLUSIONS

We found high rates of inaccurate laceration diagnosis and inappropriate application of the perineal OASI degree system, suggesting education and training are needed. Clinical studies that rely on delivery diagnosis of OASI may not be reliable.

Validation of the Use of Electronic Health Records for Classification of ADHD Status

OBJECTIVE

To validate an electronic health record (EHR)-based algorithm to classify ADHD status of pediatric patients.

METHOD

As part of an applied study, we identified all primary care patients of The Children's Hospital of Philadelphia [CHOP] health care network who were born 1987-1995 and residents of New Jersey. Patients were classified with ADHD if their EHR indicated an International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) diagnosis code of "314.x" at a clinical visit or on a list of known conditions. We manually reviewed EHRs for ADHD patients ( n = 2,030) and a random weighted sample of non-ADHD patients ( n = 807 of 13,579) to confirm the presence or absence of ADHD.

RESULTS

Depending on assumptions for inconclusive cases, sensitivity ranged from 0.96 to 0.97 (95% confidence interval [CI] = [0.95, 0.97]), specificity from 0.98 to 0.99 [0.97, 0.99], and positive predictive value from 0.83 to 0.98 [0.81, 0.99].

CONCLUSION

EHR-based diagnostic codes can accurately classify ADHD status among pediatric patients and can be used by large-scale epidemiologic and clinical studies with high sensitivity and specificity.

Etiology of Above-knee Amputations in the United States: Is Periprosthetic Joint Infection an Emerging Cause?

BACKGROUND

Above-knee amputation (AKA) is a morbid procedure and is performed for a number of conditions. Although AKA is usually performed for dysvascular disease, trauma, and malignancy, AKA is also considered in patients who have failed multiple salvage attempts at treating periprosthetic joint infection (PJI) of TKA. Although aggressive measures are being taken to treat PJI, the huge volume of TKAs might result in a large number of AKAs being performed for PJI in the United States. However, the national trends in the incidence of AKAs from different etiologies and the relative contribution of different etiologies to AKA are yet to be studied.

QUESTIONS/PURPOSES

(1) What are the temporal trends in the incidence of AKAs (from all causes) in the US population from 1998 to 2013? (2) What are the temporal trends in the incidence of AKAs by etiology (dysvascular disease, trauma, malignancy, and PJI)? (3) What are the temporal trends in the relative contribution of different etiologies to AKA?

METHODS

Using the Nationwide Inpatient Sample (NIS) from 1998 to 2013, AKAs were identified using International Classification of Diseases, 9 Revision (ICD-9) procedure code 84.17. The NIS database is the largest all-payer database in the United States containing information on approximately 20% of all the hospital admissions in the country. As a result of its sampling design, it allows for estimation of procedural volumes at the national level. All AKAs were grouped into one of the following five etiologies in a sequential manner using ICD-9 diagnosis codes: malignancy, PJI, trauma, dysvascular disease (peripheral vascular disease, diabetic, or a combination), and others. All of the numbers were converted to national estimates using sampling weights provided by the NIS, and the national incidence of AKAs resulting from various etiologies was calculated using the US population as the denominator. Poisson and linear regression analyses were used to analyze the annual trends.

RESULTS

From 1998 to 2013, the incidence of AKAs decreased by 47% from 174 to 92 AKAs per 1 million adults (incidence rate ratio [IRR]; change in the number of AKAs per 1 million adults per year; 0.96; 95% confidence interval [CI], 0.96-0.96; p < 0.001). The incidence of AKAs resulting from PJI increased by 263% (IRR, 1.07; 95% CI, 1.06-1.07; p < 0.001). An increase was also observed for AKAs from malignancy (IRR, 1.01; 95% CI, 1.00-1.02; p = 0.007), although to a smaller extent. AKAs from dysvascular causes (IRR, 0.96; 95% CI, 0.95-0.96; p < 0.001) and other etiologies (IRR, 0.97; 95% CI, 0.96-0.97; p < 0.001) decreased. There was no change in the incidence of AKAs related to trauma (IRR, 1.00; 95% CI, 0.99-1.00; p = 0.088). The proportion of AKAs resulting from PJI increased by 589% from 1998 to 2013 (coefficient = 0.18; 95% CI, 0.15-0.22; p < 0.001). The proportion of AKAs resulting from dysvascular causes decreased (coefficient = 0.18; 95% CI, 0.15-0.22; p < 0.001), whereas that resulting from malignancy (coefficient = 0.04; 95% CI, 0.03-0.05; p < 0.001) and trauma (coefficient = 0.13; 95% CI, 0.09-0.18; p < 0.001) increased.

CONCLUSIONS

The incidence of AKAs has decreased in the United States. AKAs related to dysvascular disease and other etiologies such as trauma and malignancy have either substantially decreased or remained fairly constant, whereas that resulting from PJI more than tripled. Given the increased resource utilization associated with limb loss, the results of this study suggest that national efforts to reduce disability should prioritize PJI. Further studies are required to evaluate the risk factors for AKA from PJI and to formulate better strategies to manage PJI.

LEVEL OF EVIDENCE

Level III, therapeutic study.

DPP-4 Inhibitor-Induced Rheumatoid Arthritis Among Diabetics: A Nested Case-Control Study

INTRODUCTION

The risk of rheumatoid arthritis (RA) associated with dipeptidyl peptidase-4 inhibitor (DPP-4i) use is unclear. This study assesses the RA risk associated with DPP-4i use among a diabetic cohort initiating second-line therapy.

METHODS

This was a nested case-control study, using the adult diabetic population starting second-line antidiabetic therapy from IMS LifeLink Plus^® database (2006-2015). Cases were those with two or more RA diagnosis, at least one prescription, and 180 days enrollment prior to the event date (earliest of the two: first RA diagnosis, first RA prescription). Controls were drawn from the nest after matching (1:15) with cases on index date (± 90 days), age (± 5 years), sex, and event date (imputed to have the same time difference between cohort entry and event date as the matched case). Exposure and covariate information was gathered from the 180-day period prior to event date. Conditional logistic regression was used to assess exposure among cases and controls. Adjusted analysis was carried out after controlling for important medications and comorbidities.

RESULTS

The final sample consists of 790 cases and 11,850 controls; of these, 151 cases (19.11%) and 2177 controls (18.37%) had DPP-4i claims during the exposure assessment period. DPP-4i therapy was not significantly associated with the development of RA after adjusting for covariates (OR = 1.156, 95% CI 0.936-1.429). Changing the exposure definition or exposure window to 1 year and subgroup analyses yielded similar results except for the non-insulin-using subgroup (OR = 1.299, 95% CI 1.001-1.985) which showed a significant positive association.

CONCLUSION

DPP-4i were not significantly associated with the risk of RA compared with other second-line antidiabetic therapies.

Appropriateness of ICD-coded Diagnostic Inpatient Hospital Discharge Data for Medical Practice Assessment

Objectives: We performed a systematic review to investigate the quality of diagnostic hospital discharge data (DHDD) in order to gain insight in the usefulness of these data for medical practice assessment. We investigated the methods used to evaluate data quality, factors that determine data quality and its consequences for medical practice assessment.

Methods: We selected studies in which both completeness (or sensitivity: SENS) and correctness (or positive predictive value: PPV) were measured. We used the random-effects model to calculate mean SENS and PPV and to explore the effect of a number of covariates.

Results: The 101 included studies were very heterogeneous. We distinguished six typical study designs. We found a mean SENS of 0.67 (95%CI: 0.62– 0.73) and PPV of 0.76 (95%CI: 0.73– 0.79); SENS was significantly lower for comorbidity and complication studies than for some single disease studies. PPV was significantly higher for Scandinavian countries than for other countries. Recoding compared to re-abstracting of the medical record as a gold standard gave a significantly lower PPV. Diagnostic data were considered appropriate by the authors of the studies for quality of care purposes when both SENS and PPV were at least 0.85. Only 13% of the studies fulfilled this criterion.

Conclusions: Variability in quality of care between settings can easily be overshadowed by variability in data quality. However, the use of DHDD by physicians to evaluate their own medical practice may be useful. But only if physicians are willing to critically interpret the meaning of the information for their medical practice assessment.

Administrative Databases Can Yield False Conclusions-An Example of Obesity in Total Joint Arthroplasty

BACKGROUND

Research using large administrative databases has substantially increased in recent years. Accuracy with which comorbidities are represented in these databases has been questioned. The purpose of this study was to evaluate the extent of errors in obesity coding and its impact on arthroplasty research.

METHODS

Eighteen thousand thirty primary total knee arthroplasties (TKAs) and 10,475 total hip arthroplasties (THAs) performed at a single healthcare system from 2004-2014 were included. Patients were classified as obese or nonobese using 2 methods: (1) body mass index (BMI) ≥30 kg/m² and (2) international classification of disease, 9th edition codes. Length of stay, operative time, and 90-day complications were collected. Effect of obesity on various outcomes was analyzed separately for both BMI- and coding-based obesity.

RESULTS

From 2004 to 2014, the prevalence of BMI-based obesity increased from 54% to 63% and 40% to 45% in TKA and THA, respectively. The prevalence of coding-based obesity increased from 15% to 28% and 8% to 17% in TKA and THA, respectively. Coding overestimated the growth of obesity in TKA and THA by 5.6 and 8.4 times, respectively. When obesity was defined by coding, obesity was falsely shown to be a significant risk factor for deep vein thrombosis (TKA), pulmonary embolism (THA), and longer hospital stay (TKA and THA).

CONCLUSION

The growth in obesity observed in administrative databases may be an artifact because of improvements in coding over the years. Obesity defined by coding can overestimate the actual effect of obesity on complications after arthroplasty. Therefore, studies using large databases should be interpreted with caution, especially when variables prone to coding errors are involved.

Racial Disparities in Above-knee Amputations After TKA: A National Database Study

BACKGROUND

Above-knee amputation (AKA) is a rare but devastating complication of TKA. Although racial disparities have been previously reported in the utilization of TKA, it is unclear whether disparities exist in the rates of AKA after TKA.

QUESTIONS/PURPOSES

(1) Which gender-racial group has the highest rate of AKA from septic and aseptic complications of TKA? (2) Which age groups have higher rates of AKA from septic and aseptic complications of TKA?

METHODS

Using National Inpatient Sample data from 2000 to 2011, AKAs resulting from complications of TKA were identified using a combination of International Classification of Diseases, 9th Revision procedure and diagnosis codes. Of the 341,954 AKAs identified, 9733 AKAs were the result of complications of TKA (septic complications = 8104, aseptic complications = 1629). Standardized AKA rates were calculated for different age and gender- racial groups by dividing the number of AKAs in each group with the corresponding number of TKAs. Standardized rate ratios were calculated after adjusting for demographics and comorbidities.

RESULTS

After adjusting for age and comorbidities, black men had the highest rate of AKA after TKA (adjusted rate in black men = 578 AKAs per 100,000 TKAs, standardized rate ratio [SRR] = 4.32 [confidence interval {CI}, 3.87-4.82], p < 0.001). Black men also had the highest rate of AKA after septic complications of TKA (p < 0.001). The adjusted rates of AKA were higher in patients younger than 50 years (adjusted rate = 473, SRR = 3.14 [CI, 2.94-3.36], p < 0.001) and older than 80 years (adjusted rate = 297, SRR = 1.85 [CI, 1.76-1.95], p < 0.001).

CONCLUSIONS

The rising demand for TKA has led to an increase in the number of AKAs performed for complications of TKA in the United States. Although we did not find an increase in the rate of AKA during the study period, certain populations, including black men and patients older than 80 years and younger than 50 years, had higher rates of AKA. Further studies are required to understand the reasons for these disparities and measures should be undertaken to eliminate these disparities.

LEVEL OF EVIDENCE

Level III, therapeutic study.

Administrative Algorithms to identify Avascular necrosis of bone among patients undergoing upper or lower extremity magnetic resonance imaging: a validation study

BACKGROUND

Studies of the epidemiology and outcomes of avascular necrosis (AVN) require accurate case-finding methods. The aim of this study was to evaluate performance characteristics of a claims-based algorithm designed to identify AVN cases in administrative data.

METHODS

Using a centralized patient registry from a US academic medical center, we identified all adults aged ≥18 years who underwent magnetic resonance imaging (MRI) of an upper/lower extremity joint during the 1.5 year study period. A radiologist report confirming AVN on MRI served as the gold standard. We examined the sensitivity, specificity, positive predictive value (PPV) and positive likelihood ratio (LR⁺) of four algorithms (A-D) using International Classification of Diseases, 9th edition (ICD-9) codes for AVN. The algorithms ranged from least stringent (Algorithm A, requiring ≥1 ICD-9 code for AVN [733.4X]) to most stringent (Algorithm D, requiring ≥3 ICD-9 codes, each at least 30 days apart).

RESULTS

Among 8200 patients who underwent MRI, 83 (1.0% [95% CI 0.78-1.22]) had AVN by gold standard. Algorithm A yielded the highest sensitivity (81.9%, 95% CI 72.0-89.5), with PPV of 66.0% (95% CI 56.0-75.1). The PPV of algorithm D increased to 82.2% (95% CI 67.9-92.0), although sensitivity decreased to 44.6% (95% CI 33.7-55.9). All four algorithms had specificities >99%.

CONCLUSION

An algorithm that uses a single billing code to screen for AVN among those who had MRI has the highest sensitivity and is best suited for studies in which further medical record review confirming AVN is feasible. Algorithms using multiple billing codes are recommended for use in administrative databases when further AVN validation is not feasible.

Early Hospital Readmission and Mortality Risk after Surgical Treatment of Proximal Humerus Fractures in a Community-Based Health Care Organization

CONTEXT

Surgical treatment for proximal humerus fractures has increased exponentially. Recent health care policies incentivize centers to reduce hospital readmission rates. Better understanding of risk factors for readmission and early mortality in this population will assist in identifying favorable risk-benefit patient profiles.

OBJECTIVE

To identify incidence and risk factors of 30-day hospital readmission rate and 1-year mortality rate after open surgery of proximal humerus fractures.

DESIGN

Retrospective cohort analysis from Kaiser Permanente Southern California Region database.

METHODS

Using International Classification of Diseases, Ninth Revision, diagnosis and procedure codes, all operative proximal humerus fractures were validated. Hospital readmission, one-year mortality, and demographic and medical data were collected. A logistic regression test was performed to assess potential risk factors for outcomes.

RESULTS

From 1387 surgical patients, the 30-day all-cause readmission rate was 5.6%. Forty percent of hospital read-missions were due to surgery-related reasons. Severe liver disease (odds ratio [OR], 3.48, 95% confidence interval [CI] = 1.42-8.55) and LACE (length of stay, acuity of admission, comorbidities, and number of Emergency Department visits in the previous 6 months) index score ≥ 10 (OR, 4.47, 95% CI = 2.54-7.86) were independent risk factors of readmission on multivariate analysis. The 1-year mortality rate was 4.86%. Multivariate analysis showed length of hospital stay (OR 1.11, 95% CI = 1.05-1.19), cancer (OR 3.38, 95% CI = 1.61-7.10), 30-day readmission (OR 3.31, 95% CI = 1.34-8.21), and Charlson comorbidity index greater than or equal to 4 (OR 13.94, 95% CI = 4.40-44.17) predicted higher mortality risk.

CONCLUSION

After open treatment of proximal humerus fractures, there was a 5.6% all-cause 30-day hospital readmission rate. Surgical complications accounted for 40% of read-missions. Severe liver disease and LACE score correlated best with postoperative 30-day readmission risk. Length of hospital stay, preexisting cancer, 30-day readmission, and Charlson comorbidity index were predictive of 1-year mortality.

The use of administrative health care databases to identify patients with rheumatoid arthritis

OBJECTIVE

To validate and compare the decision rules to identify rheumatoid arthritis (RA) in administrative databases.

METHODS

A study was performed using administrative health care data from a population of 1 million people who had access to universal health care. Information was available on hospital discharge abstracts and physician billings. RA cases in health administrative databases were matched 1:4 by age and sex to randomly selected controls without inflammatory arthritis. Seven case definitions were applied to identify RA cases in the health administrative data, and their performance was compared with the diagnosis by a rheumatologist. The validation study was conducted on a sample of individuals with administrative data who received a rheumatologist consultation at the Arthritis Center of Nova Scotia.

RESULTS

We identified 535 RA cases and 2,140 non-RA, noninflammatory arthritis controls. Using the rheumatologist's diagnosis as the gold standard, the overall accuracy of the case definitions for RA cases varied between 68.9% and 82.9% with a kappa statistic between 0.26 and 0.53. The sensitivity and specificity varied from 20.7% to 94.8% and 62.5% to 98.5%, respectively. In a reference population of 1 million, the estimated annual number of incident cases of RA was between 176 and 1,610 and the annual number of prevalent cases was between 1,384 and 5,722.

CONCLUSION

The accuracy of case definitions for the identification of RA cases from rheumatology clinics using administrative health care databases is variable when compared to a rheumatologist's assessment. This should be considered when comparing results across studies. This variability may also be used as an advantage in different study designs, depending on the relative importance of sensitivity and specificity for identifying the population of interest to the research question.

Concordance Between Clinical Diagnosis and Medicare Claims of Depression Among Older Primary Care Patients

OBJECTIVE

To identify patient characteristics associated with concordance of Medicare claims with clinically identified depression.

METHODS

The authors studied a cohort of 742 older primary care patients linked to Medicare claims data using the Structured Clinical Interview for Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition major depressive disorder and clinically significant minor depression.

RESULTS

Among 474 patients with depression, 198 patients had a Medicare claim for depression (sensitivity: 42%; 95% confidence interval [CI]: 37%-46%). Among 268 patients who did not meet criteria for depression, 235 patients did not have a Medicare claim for depression (specificity: 88%; 95% CI: 83%-91%). After adjustment for demographic and clinical characteristics, non-white participants were nearly twice as likely not to have Medicare claims for depression among patients who met criteria for depression ("false negatives"). Smoking status, depression severity (Hamilton Depression Rating Scale), cardiovascular disease, and more primary care physician office visits were also significantly associated with decreased odds to be false negatives. In contrast, after covariate adjustment, white race and chronic pulmonary disease were associated with increased odds of a Medicare claim for depression among patients who did not meet criteria for depression ("false positives"). Using weights based on the screened sample, the positive predictive value of a Medicare claim for depression was 66% (95% CI [63%, 69%]), whereas the negative predictive value was 77% (95% CI [76%, 78%]).

CONCLUSION

Investigators using Medicare data to study depression must recognize that diagnoses of depression from Medicare data may be biased by patient ethnicity and the presence of medical comorbidity.

Health care resource utilization and costs associated with nonfatal major adverse cardiovascular events

BACKGROUND

Risk of recurrent cardiovascular events following an initial cardiovascular-related hospitalization remains high despite available interventions. Rates of cardiovascular events as well as associated health care resource utilization and costs are needed to assess the value of treatments.

OBJECTIVE

To quantify, in patients with previous hospitalization for acute coronary syndrome (ACS), rates of nonfatal major adverse cardiovascular events (MACE) and secondary coronary events (SCE), as well as health care utilization and costs associated with a first MACE.

METHODS

Administrative data from a large population of commercial managed care and managed Medicare enrollees in the United States were retrospectively analyzed. Patients with an ACS-related hospitalization from 2006 to 2011 were followed for 12 months to assess subsequent MACE and SCE rates. Patients were aged ≥ 18 years at initial ACS hospitalization (the index episode) and had ≥ 12 months of continuous health plan enrollment before and after the end of the index episode. Resource utilization and costs during a first MACE were assessed. Multivariable analyses were used to assess the associations between cardiovascular risk factors and the occurrence of a MACE, as well as the costs incurred during a first MACE.

RESULTS

Of 75,231 study patients identified, 3.3% had a MACE and 8.3% had an SCE during the 12-month follow-up. Median time to first MACE and SCE from end of the index episode was 4.6 and 3.7 months, respectively. Mean MACE-related cost incurred during the first MACE was $19,642. Logistic analyses showed that age and diabetes were associated with increased odds of a MACE, while index ACS episodes involving ST-elevation myocardial infarction were associated with reduced odds. Findings from generalized linear models indicated that statin use and age were associated with lower episode-related costs and that MACE occurrence within 3 months of ACS hospitalization was associated with increased episode-related costs.

CONCLUSIONS

MACEs and SCEs represent a common and costly burden in the year following ACS hospitalization. Our findings may inform future economic assessments of new therapies aimed at prevention of MACEs and SCEs.

Validation of Diagnostic Groups Based on Health Care Utilization Data Should Adjust for Sampling Strategy

Supplemental Digital Content is available in the text.

Objective:

Valid measurement of outcomes such as disease prevalence using health care utilization data is fundamental to the implementation of a “learning health system.” Definitions of such outcomes can be complex, based on multiple diagnostic codes. The literature on validating such data demonstrates a lack of awareness of the need for a stratified sampling design and corresponding statistical methods. We propose a method for validating the measurement of diagnostic groups that have: (1) different prevalences of diagnostic codes within the group; and (2) low prevalence.

Methods:

We describe an estimation method whereby: (1) low-prevalence diagnostic codes are oversampled, and the positive predictive value (PPV) of the diagnostic group is estimated as a weighted average of the PPV of each diagnostic code; and (2) claims that fall within a low-prevalence diagnostic group are oversampled relative to claims that are not, and bias-adjusted estimators of sensitivity and specificity are generated.

Application:

We illustrate our proposed method using an example from population health surveillance in which diagnostic groups are applied to physician claims to identify cases of acute respiratory illness.

Conclusions:

Failure to account for the prevalence of each diagnostic code within a diagnostic group leads to the underestimation of the PPV, because low-prevalence diagnostic codes are more likely to be false positives. Failure to adjust for oversampling of claims that fall within the low-prevalence diagnostic group relative to those that do not leads to the overestimation of sensitivity and underestimation of specificity.

A validation study of a new classification algorithm to identify rheumatoid arthritis using administrative health databases: case-control and cohort diagnostic accuracy studies. Results from the RECord linkage On Rheumatic Diseases study of the Italian Society for Rheumatology

OBJECTIVES

To develop and validate a new algorithm to identify patients with rheumatoid arthritis (RA) and estimate disease prevalence using administrative health databases (AHDs) of the Italian Lombardy region.

DESIGN

Case-control and cohort diagnostic accuracy study.

METHODS

In a randomly selected sample of 827 patients drawn from a tertiary rheumatology centre (training set), clinically validated diagnoses were linked to administrative data including diagnostic codes and drug prescriptions. An algorithm in steps of decreasing specificity was developed and its accuracy assessed calculating sensitivity/specificity, positive predictive value (PPV)/negative predictive value, with corresponding CIs. The algorithm was applied to two validating sets: 106 patients from a secondary rheumatology centre and 6087 participants from the primary care. Alternative algorithms were developed to increase PPV at population level. Crude and adjusted prevalence estimates taking into account algorithm misclassification rates were obtained for the Lombardy region.

RESULTS

The algorithms included: RA certification by a rheumatologist, certification for other autoimmune diseases by specialists, RA code in the hospital discharge form, prescription of disease-modifying antirheumatic drugs and oral glucocorticoids. In the training set, a four-step algorithm identified clinically diagnosed RA cases with a sensitivity of 96.3 (95% CI 93.6 to 98.2) and a specificity of 90.3 (87.4 to 92.7). Both external validations showed highly consistent results. More specific algorithms achieved >80% PPV at the population level. The crude RA prevalence in Lombardy was 0.52%, and estimates adjusted for misclassification ranged from 0.31% (95% CI 0.14% to 0.42%) to 0.37% (0.25% to 0.47%).

CONCLUSIONS

AHDs are valuable tools for the identification of RA cases at the population level, and allow estimation of disease prevalence and to select retrospective cohorts.

Variation in the Use of Therapy following Distal Radius Fractures in the United States

Background:

Distal radius fractures (DRFs) are one of the most common injuries among the elderly, resulting in significant expense and disability. The specific aims of this study are (1) to examine rates of therapy following DRFs and (2) to identify those factors that influence utilization of therapy and time span between DRF treatment and therapy among a national cohort of elderly patients.

Methods:

We examined national use of physical and occupational therapy among all Medicare beneficiaries who suffered DRFs between January 1, 2007, and October 1, 2007, and assessed the effect of treatment, patient-related, and surgeon-related factors on utilization of therapy.

Results:

Overall, 20.6% of patients received either physical or occupational therapy following DRF. Use of therapy varied by DRF treatment, and patients who underwent open reduction and internal fixation were more likely to receive therapy compared with patients who received closed reduction. Patients who received open reduction and internal fixation were also referred to therapy earlier compared with patients who received external fixation, percutaneous pinning, and closed reduction. Surgeon specialization is associated with greater use of postoperative therapy. Patient predictors of therapy use include younger age, female sex, higher socioeconomic status, and fewer comorbidity conditions.

Conclusion:

Use of therapy following DRF varies significantly by both patient- and surgeon-related factors. Identifying patients who benefit from postinjury therapy can allow for better resource utilization following these common injuries.

US trends in rates of arthroplasty for inflammatory arthritis including rheumatoid arthritis, juvenile idiopathic arthritis, and spondyloarthritis

OBJECTIVE

Although rates of arthroplasty have increased dramatically, rates among patients with rheumatoid arthritis (RA) are reported to be decreasing. It is not known if this is also the case among patients with other inflammatory arthritides. This study was undertaken to evaluate rates of arthroplasty due to RA, juvenile idiopathic arthritis (JIA), spondyloarthritis (SpA), and a composite group of patients with inflammatory arthritides (IA), compared to arthroplasty rates among patients without inflammatory or autoimmune conditions.

METHODS

Administrative discharge databases (State Inpatient Databases of the Healthcare Cost and Utilization Project, New York Department of Health Statewide Planning and Research Cooperative System, California Statewide Health Planning and Development) were used to compare rates of knee, hip, and shoulder arthroplasty occurring from 1991 to 2005.

RESULTS

Of 2,839,325 arthroplasties in 1991-2005, 2.7% were performed in patients with IA. The rate of arthroplasty for noninflammatory conditions doubled (124.5 per 100,000 persons in 1991 versus 247.5 per 100,000 persons in 2005), while the rate for IA remained stable at 5.1 per 100,000. Rates of arthroplasty for RA decreased slightly (4.6 per 100,000 versus 4.5 per 100,000) and those for JIA decreased by nearly 50% (0.22 per 100,000 versus 0.13 per 100,000), but the rate of arthroplasty for SpA increased by nearly 50% (0.22 per 100,000 versus 0.31 per 100,000). Age at the time of arthroplasty increased for patients with RA (mean ± SD 63.4 ± 12.7 years versus 64.9 ± 12.8 years), JIA (30.9 ± 12.2 years versus 36.7 ± 14.9 years), and SpA (54.3 ± 16.1 years versus 60.4 ± 13.9 years). However, the mean age at the time of arthroplasty among non-IA cases decreased (71.5 ± 11.8 years versus 69.0 ± 12.0 years).

CONCLUSION

This population-based study is the first to show that arthroplasty rates have decreased significantly among patients with JIA and minimally among patients with RA, and have increased among patients with SpA. The increased age at the time of arthroplasty among patients with JIA and SpA suggests that these patients are increasingly able to defer surgical interventions. Further research is needed to assess the ongoing effect of biologic agents on the need for arthroplasties in patients with IA.

A systematic review of validated methods for identifying patients with rheumatoid arthritis using administrative or claims data

PURPOSE

To review the evidence supporting the validity of billing, procedural, or diagnosis code, or pharmacy claim-based algorithms used to identify patients with rheumatoid arthritis (RA) in administrative and claim databases.

METHODS

We searched the MEDLINE database from 1991 to September 2012 using controlled vocabulary and key terms related to RA and reference lists of included studies were searched. Two investigators independently assessed the full text of studies against pre-determined inclusion criteria and extracted the data. Data collected included participant and algorithm characteristics.

RESULTS

Nine studies reported validation of computer algorithms based on International Classification of Diseases (ICD) codes with or without free-text, medication use, laboratory data and the need for a diagnosis by a rheumatologist. These studies yielded positive predictive values (PPV) ranging from 34 to 97% to identify patients with RA. Higher PPVs were obtained with the use of at least two ICD and/or procedure codes (ICD-9 code 714 and others), the requirement of a prescription of a medication used to treat RA, or requirement of participation of a rheumatologist in patient care. For example, the PPV increased from 66 to 97% when the use of disease-modifying antirheumatic drugs and the presence of a positive rheumatoid factor were required.

CONCLUSIONS

There have been substantial efforts to propose and validate algorithms to identify patients with RA in automated databases. Algorithms that include more than one code and incorporate medications or laboratory data and/or required a diagnosis by a rheumatologist may increase the PPV.

Treatment patterns in patients with metastatic melanoma: a retrospective analysis

Objective. To describe treatment patterns and factors influencing treatment in a real-world setting of US patients with metastatic melanoma (MM). Methods. This was a retrospective claims-based study among patients with MM diagnosed between 2005 and 2010 identified from MarketScan databases. Results. Of 2546 MM patients, 66.8% received surgery, 44.7% received radiation, 38.7% received systemic therapies, and 17.7% received all modalities. Patients with lung, brain, liver, or bone metastases were less likely to undergo surgery (all P < 0.0001); patients with lung (P = 0.04), brain (P < 0.001), or liver metastases (P = 0.03) were more likely to receive systemic therapies; patients with brain (P < 0.0001) or bone metastases (P < 0.0001) were more likely to receive radiation therapy. Oncologists were more likely to recommend systemic therapy (P < 0.0001) or radiation (P < 0.0001), while dermatologists were more likely to recommend surgery (P = 0.002). Monotherapy was the dominant systemic therapy (82.4% patients as first-line). Conclusions. Only 39% of MM patients received systemic therapies, perhaps reflecting efficacy and safety limitations of conventional systemic therapies for MM. Among those receiving systemic therapy, monotherapy was the most common approach. Sites of metastases and physician speciality influenced treatment patterns. This study serves as a baseline against which future treatment pattern studies, following approval of new agents, can be compared.

Identifying diabetics in Medicare claims and survey data: implications for health services research

Background

Diabetes health services research often utilizes secondary data sources, including survey self-report and Medicare claims, to identify and study the diabetic population, but disagreement exists between these two data sources. We assessed agreement between the Chronic Condition Warehouse diabetes algorithm for Medicare claims and self-report measures of diabetes. Differences in healthcare utilization outcomes under each diabetes definition were also explored.

Methods

Claims data from the Medicare Beneficiary Annual Summary File were linked to survey and blood data collected from the 2006 Health and Retirement Study. A Hemoglobin A1c reading, collected on 2,028 respondents, was used to reconcile discrepancies between the self-report and Medicare claims measures of diabetes. T-tests were used to assess differences in healthcare utilization outcomes for each diabetes measure.

Results

The Chronic Condition Warehouse (CCW) algorithm yielded a higher rate of diabetes than respondent self-reports (27.3 vs. 21.2, p < 0.05). A1c levels of discordant claims-based diabetics suggest that these patients are not diabetic, however, they have high rates of healthcare spending and utilization similar to diabetics.

Conclusions

Concordance between A1c and self-reports was higher than for A1c and the CCW algorithm. Accuracy of self-reports was superior to the CCW algorithm. False positives in the claims data have similar utilization profiles to diabetics, suggesting minimal bias in some types of claims-based analyses, though researchers should consider sensitivity analysis across definitions for health services research.

Variation in the incidence of distal radius fractures in the U.S. elderly as related to slippery weather conditions

BACKGROUND

Distal radius fractures are costly and debilitating injuries, especially for the elderly. These fractures often occur from falls and commonly occur outdoors. Inclement weather may increase the risk of fall-related injuries. Small studies have reported an increased risk of distal radius fracture caused by inclement winter weather; larger studies are lacking.

METHODS

The authors analyzed a sample of 2007 Medicare claims for distal radius fracture. Weather data were collected for the date and location of each distal radius fracture in the authors' analysis cohort. A novel slipperiness score was used as a measure of the severity of slippery outdoor conditions. Negative binomial regression models evaluated the correlation between slipperiness and distal radius fracture occurrence.

RESULTS

Risk of distal radius fracture was higher in winter months (incidence rate ratio, 1.2; p < 0.001). Days with average temperature less than or equal to 32°F (incidence rate ratio, 1.36; p < 0.001), snow/ice on the ground at the start of the day (incidence rate ratio, 1.45; p < 0.001), and freezing rain (incidence rate ratio, 1.24; p = 0.025) all had an increased risk of distal radius fracture. The risk of sustaining a distal radius fracture was increased 21 percent on days with a slipperiness score above 4 (incidence rate ratio, 1.21; p = 0.007). For each increase in slipperiness score above 4, the incidence rate ratio of distal radius fracture increased as well.

CONCLUSIONS

Weather events that create slippery walking conditions, often in the winter, result in an increased risk of distal radius fracture in the elderly. This finding can be used to support resource allocation and awareness and prevention campaigns.

CLINICAL QUESTION/LEVEL OF EVIDENCE

Risk, IV.

Mining high-dimensional administrative claims data to predict early hospital readmissions

BACKGROUND

Current readmission models use administrative data supplemented with clinical information. However, the majority of these result in poor predictive performance (area under the curve (AUC)<0.70).

OBJECTIVE

To develop an administrative claim-based algorithm to predict 30-day readmission using standardized billing codes and basic admission characteristics available before discharge.

MATERIALS AND METHODS

The algorithm works by exploiting high-dimensional information in administrative claims data and automatically selecting empirical risk factors. We applied the algorithm to index admissions in two types of hospitalized patient: (1) medical patients and (2) patients with chronic pancreatitis (CP). We trained the models on 26,091 medical admissions and 3218 CP admissions from The Johns Hopkins Hospital (a tertiary research medical center) and tested them on 16,194 medical admissions and 706 CP admissions from Johns Hopkins Bayview Medical Center (a hospital that serves a more general patient population), and vice versa. Performance metrics included AUC, sensitivity, specificity, positive predictive values, negative predictive values, and F-measure.

RESULTS

From a pool of up to 5665 International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) diagnoses, 599 ICD-9-CM procedures, and 1815 Current Procedural Terminology codes observed, the algorithm learned a model consisting of 18 attributes from the medical patient cohort and five attributes from the CP cohort. Within-site and across-site validations had an AUC≥0.75 for the medical patient cohort and an AUC≥0.65 for the CP cohort.

CONCLUSIONS

We have created an algorithm that is widely applicable to various patient cohorts and portable across institutions. The algorithm performed similarly to state-of-the-art readmission models that require clinical data.

Systemic autoimmune rheumatic disease prevalence in Canada: updated analyses across 7 provinces

OBJECTIVE

To estimate systemic autoimmune rheumatic disease (SARD) prevalence across 7 Canadian provinces using population-based administrative data evaluating both regional variations and the effects of age and sex.

METHODS

Using provincial physician billing and hospitalization data, cases of SARD (systemic lupus erythematosus, scleroderma, primary Sjögren syndrome, polymyositis/dermatomyositis) were ascertained. Three case definitions (rheumatology billing, 2-code physician billing, and hospital diagnosis) were combined to derive a SARD prevalence estimate for each province, categorized by age, sex, and rural/urban status. A hierarchical Bayesian latent class regression model was fit to account for the imperfect sensitivity and specificity of each case definition. The model also provided sensitivity estimates of different case definition approaches.

RESULTS

Prevalence estimates for overall SARD ranged between 2 and 5 cases per 1000 residents across provinces. Similar demographic trends were evident across provinces, with greater prevalence in women and in persons over 45 years old. SARD prevalence in women over 45 was close to 1%. Overall sensitivity was poor, but estimates for each of the 3 case definitions improved within older populations and were slightly higher for men compared to women.

CONCLUSION

Our results are consistent with previous estimates and other North American findings, and provide results from coast to coast, as well as useful information about the degree of regional and demographic variations that can be seen within a single country. Our work demonstrates the usefulness of using multiple data sources, adjusting for the error in each, and providing estimates of the sensitivity of different case definition approaches.

Cost offsets to medicare attributable to receipt of hip, knee, and shoulder arthroplasty

OBJECTIVE

To estimate trends in numbers of and Medicare payments for hip, knee, and shoulder arthroplasties for beneficiaries with osteoarthritis (OA) and potential savings to Medicare from arthroplasty during followup.

METHODS

The analysis was based on longitudinal 5% Medicare enrollment and claims data for 1992-2010. The analysis of changes in Medicare payments attributable to total arthroplasty receipt used propensity score matching to obtain beneficiary control groups matched on demographic characteristics, general health, joint pain, and Medicare payments by major condition in the year preceding the index arthroplasty. An average treatment effect on the treated (ATT) overall and for each major condition was calculated for payments for care 7-36 months following the index arthroplasty procedure.

RESULTS

Growth in incident OA diagnoses of the hip, knee, and shoulder was substantially higher than growth in real Medicare spending on hip, knee, and shoulder arthroplasties. ATTs showed a mean saving to Medicare of $471/beneficiary/procedure for hip, no difference for knee, and a payment increase of $1,062 for shoulder arthroplasty during followup. For hip arthroplasty, the largest savings was for the circulatory system. For shoulder arthroplasty, increased payments during followup reflected increased payments for musculoskeletal care, especially for hip and knee arthroplasty. Overall, payment differences during followup by major condition were small.

CONCLUSIONS

Provision of hip but not knee and shoulder arthroplasty generated savings to Medicare during followup, but even for hip arthroplasty, the cost offset during followup was small relative to the program cost for the procedure itself.

Residential exposure to aircraft noise and hospital admissions for cardiovascular diseases: multi-airport retrospective study

OBJECTIVE

To investigate whether exposure to aircraft noise increases the risk of hospitalization for cardiovascular diseases in older people (≥ 65 years) residing near airports.

DESIGN

Multi-airport retrospective study of approximately 6 million older people residing near airports in the United States. We superimposed contours of aircraft noise levels (in decibels, dB) for 89 airports for 2009 provided by the US Federal Aviation Administration on census block resolution population data to construct two exposure metrics applicable to zip code resolution health insurance data: population weighted noise within each zip code, and 90th centile of noise among populated census blocks within each zip code.

SETTING

2218 zip codes surrounding 89 airports in the contiguous states.

PARTICIPANTS

6 027 363 people eligible to participate in the national medical insurance (Medicare) program (aged ≥ 65 years) residing near airports in 2009.

MAIN OUTCOME MEASURES

Percentage increase in the hospitalization admission rate for cardiovascular disease associated with a 10 dB increase in aircraft noise, for each airport and on average across airports adjusted by individual level characteristics (age, sex, race), zip code level socioeconomic status and demographics, zip code level air pollution (fine particulate matter and ozone), and roadway density.

RESULTS

Averaged across all airports and using the 90th centile noise exposure metric, a zip code with 10 dB higher noise exposure had a 3.5% higher (95% confidence interval 0.2% to 7.0%) cardiovascular hospital admission rate, after controlling for covariates.

CONCLUSIONS

Despite limitations related to potential misclassification of exposure, we found a statistically significant association between exposure to aircraft noise and risk of hospitalization for cardiovascular diseases among older people living near airports.

Development of an algorithm for identifying rheumatoid arthritis in the Korean National Health Insurance claims database

This study aimed to develop an identification algorithm for validating the International Classification of Diseases-Tenth diagnostic codes for rheumatoid arthritis (RA) in the Korean National Health Insurance (NHI) claims database. An individual copayment beneficiaries program for rare and intractable diseases, including seropositive RA (M05), began in South Korea in July 2009. Patients registered in this system pay only 10 % of their total medical costs, but registration requires an official report from a doctor documenting that the patient fulfills the 1987 ACR criteria. We regarded patients registered in this system as gold standard RA and examined the validity of several algorithms to define RA diagnosis using diagnostic codes and prescription data. We constructed nine algorithms using two highly specific prescriptions (positive predictive value >90 % and specificity >90 %) and one prescription with high sensitivity (>80 %) and accuracy (>75 %). A total of 59,823 RA patients were included in this validation study. Among them, 50,082 (83.7 %) were registered in the individual copayment beneficiaries program and considered true RA. We tested nine algorithms that incorporated two specific regimens [biologics and leflunomide alone, methotrexate plus leflunomide, or more than 3 disease-modifying anti-rheumatic drugs (DMARDs)] and one sensitive drug (any non-steroidal anti-inflammatory drug (NSAID), any DMARD, or any NSAID plus any DMARD). The algorithm that included biologics, more than 3 DMARDs, and any DMARD yielded the highest accuracy (91.4 %). Patients with RA diagnostic codes with prescription of biologics or any DMARD can be considered as accurate cases of RA in Korean NHI claims database.

Longer Term Effects of Total Knee Arthroplasty From a National Longitudinal Study

Objective: This study used data for 1996-2010 from a U.S. longitudinal sample of elderly individuals from the Health and Retirement Study (HRS) merged with Medicare claims data to assess changes in several dimensions of physical functioning and general health up to 68 months following total knee arthroplasty (TKA) receipt. Method: Using propensity score matching, we assessed outcomes at follow-up for Medicare beneficiaries receiving TKA and a comparable group of beneficiaries with the same osteoarthritis diagnoses (controls). Results: Receipt of TKA was most often associated with improvements in physical functioning, especially in physical functioning measures most directly related to the knee. General health of TKA recipients only improved relative to controls on 1 of the 3 study general health measures. Discussion: Improvements in physical functioning of TKA recipients persisted in this longer term analysis of outcome in a nationally representative population study.

Identification of rheumatoid arthritis patients using an administrative database: a Veterans Affairs study

OBJECTIVE

The accuracy of the diagnosis is vital when administrative databases are used for pharmacoepidemiologic and outcome studies. Data pertaining to the utility of databases for rheumatoid arthritis (RA) are sparse and variable. We assessed the utility of various diagnostic algorithms to identify RA patients within the Veterans Health Administration (VHA) databases.

METHODS

Using the International Classification of Diseases, Ninth Revision code for RA at 2 visits at least 6 months apart, we identified 1,779 patients between October 1, 1998 and September 30, 2009 in our local Veterans Affairs Medical Center (VAMC) administrative database. Disease-modifying antirheumatic drug (DMARD) use was ascertained from the pharmacy database. Cases were analyzed based on DMARD therapy and RA codes at clinic visits. A total of 543 patients' medical records, selected by stratification and random selection on the basis of their visits, were reviewed to ascertain the clinicians' diagnoses and clinical criteria documentation. Positive predictive values (PPVs) were calculated for various database case identification algorithms using diagnosis of RA by medical record review as the gold standard.

RESULTS

The PPV for identification of RA with 2 RA codes 6 months apart was 30.9%. Addition of DMARD therapy increased the PPV to 60.4%. The PPV further increased to 91.4% when having an RA code at the last VAMC rheumatology clinic visit criterion was added. An algorithm using only 2 administrative RA codes 6 months apart had a low PPV for correctly identifying patients with RA in the VHA database.

CONCLUSION

Including DMARD therapy and requiring an RA code at the last visit with a rheumatologist increased the performance of the data extraction algorithm.

Cost-related medication nonadherence in older patients with rheumatoid arthritis

OBJECTIVE

Economic access to costly medications including biologic agents can be challenging. Our objective was to examine whether patients with rheumatoid arthritis (RA) are at particular risk for cost-related medication nonadherence (CRN) and spending less on basic needs.

METHODS

We identified a nationally representative sample of older adults with RA (n = 1100) in the Medicare Current Beneficiary Survey (2004-2008) and compared them to older adults with other morbidities categorized by chronic disease count: 0 (n = 5898), 1-2 (n = 30,538), and ≥ 3 (n = 34,837). We compared annual rates of self-reported CRN (skipping or reducing medication doses or not obtaining prescriptions because of cost) as well as spending less on basic needs to afford medications and tested for differences using survey-weighted logistic regression analyses adjusted for demographic characteristics, health status, and prescription drug coverage.

RESULTS

In the RA sample, the unadjusted weighted prevalence of CRN ranged from 20.7% in 2004 to 18.4% in 2008 as compared to 18.5% and 11.9%, respectively, in patients with 3 or more non-RA conditions. In adjusted analyses, having RA was associated with a 3.5-fold increase in the risk of CRN (OR 3.52, 95% CI 2.63-4.71) and almost a 2.5-fold risk of spending less on basic needs (OR 2.41, 95% CI 1.78-3.25) as compared to those without a chronic condition.

CONCLUSION

Patients with RA experience a high prevalence of CRN and forgoing of basic needs, more than do older adults with multiple other chronic conditions. The situation did not improve during a period of policy change aimed at alleviating high drug costs.

Association between oral 5-ASA adherence and health care utilization and costs among patients with active ulcerative colitis

BACKGROUND

Observational cohort study to assess the association between adherence to oral 5-aminosalicylates (5-ASAs) and all-cause costs and health care utilization among patients with active ulcerative colitis (UC) in the United States.

METHODS

Retrospective analysis of insurance claims from June 1997 to August 2006 in the LifeLink Database. Patient criteria: aged 18 or older with one or more claim(s) between June 1997 and August 2005 for a UC diagnosis and at least one oral 5-ASA prescription on or after the first observed UC diagnosis; continuous enrollment for at least 6 months prior to and 12 months following 5-ASA initiation (index date). As a proxy for active disease, patients needed to have at least two UC-specific non-pharmacy claims, at least 30 days of 5-ASA treatment and at least one corticosteroid prescription within the 12-month post-index period. Cumulative exposure to oral 5-ASAs over the 12-month period was calculated using the medication possession ratio (MPR). Patients with an MPR of at least 0.80 were classified as adherent. All-cause medical and pharmacy resource utilization and costs were computed over the 12-month post-index period and compared between adherent and nonadherent patients.

RESULTS

1,693 UC patients met study inclusion criteria: 72% were nonadherent to 5-ASA treatment (n=1,217) and 28% were adherent (n=476) in the 12-month study period. Compared with nonadherent patients, adherent patients had 31% fewer hospitalizations (P=0.0025) and 34% fewer emergency department admissions (P=0.0016). Adherent patients had 25% more pharmacy prescriptions overall (P <0.0001) and 71% more UC-related pharmacy prescriptions (P <0.0001) than did nonadherent patients. Total all-cause health care utilization was 1.13 times higher for adherent patients than for nonadherent patients (P=0.0002). After adjusting for covariates, total all-cause costs were 29% higher for nonadherent patients than for adherent patients (mean [95% confidence interval]: $13,465 [$13,094, $13,835] vs $17,339 [$17,033, $17,645]).

CONCLUSIONS

Approximately three-quarters of patients with active UC were not adherent with their prescribed doses of oral 5-ASA. Nonadherence was associated with higher total all-cause costs. The key driver of decreased costs among adherent patients was inpatient hospitalizations, which more than offset these patients' expected higher pharmacy costs.

Time trends in medication use and expenditures in older patients with rheumatoid arthritis

BACKGROUND

We sought to examine how expansions in insurance coverage of nonbiologic and biologic disease-modifying antirheumatic drugs affected the access, costs, and health status of older patients with rheumatoid arthritis.

METHODS

We identified a nationally representative sample of older adults with rheumatoid arthritis in the 2000-2006 Medicare Current Beneficiary Survey (unweighted n=1051). We examined changes in disease-modifying antirheumatic drug use, self-reported health status, functional status (activities of daily living), and total costs and out-of-pocket costs for medical care and prescription drugs. Tests for time trends were conducted using weighted regressions.

RESULTS

Between 2000 and 2006, the proportion of older adults with rheumatoid arthritis who received biologics tripled (4.6% vs 13.2%, P=.01), whereas the proportion of people who used a nonbiologic did not change. During the same period, the proportion of older patients with rheumatoid arthritis rating their health as excellent/good significantly increased (43.0% in 2000 to 55.6% in 2006; P=.015). Significant improvements occurred in activities of daily living measures of functional status. Total prescription drug costs (in 2006 US dollars) increased from $2645 in 2000 to $4685 in 2006, P=.0001, whereas out-of-pocket prescription costs remained constant ($842 in 2000 vs $832 in 2006; P=.68). Total medical costs did not significantly increase ($16,563 in 2000 vs $19,510 in 2006; P=.07).

CONCLUSION

Receipt of biologics in older adults with rheumatoid arthritis increased over a period of time when insurance coverage was expanded without increasing patients' out-of-pocket costs. During this time period, concurrent improvements in self-reported health status and functional status suggest improved arthritis care.

Trends in incidence and outcomes of revision total hip arthroplasty in Spain: a population based study

Background

To analyze changes in incidence and outcomes of patients undergoing revision total hip arthroplasty (RTHA) over an 8-year study period in Spain.

Methods

We selected all surgical admissions in individuals aged ≥ 40 years who underwent RTHA (ICD-9-CM procedure code 81.53) between 2001 and 2008 from the Spanish National Hospital Discharge Database. Age- and sex-specific incidence rates, Charlson co-morbidity index, length of stay (LOS), costs and in-hospital mortality (IHM) were estimated for each year. Multivariate analyses were conducted to asses time trends.

Results

32, 280 discharges of patients (13, 391 men/18, 889 women) having undergone RTHA were identified. Overall crude incidence showed a small but significant increase from 20.2 to 21.8 RTHA per 100, 000 inhabitants from 2001 to 2008 (p < 0.01).

The incidence increased for men (17.7 to 19.8 in 2008) but did not vary for women (22.3 in 2001 and 22.2 in 2008). Greater increments were observed in patients older than 84 years and in the age group 75-84. In 2001, 19% of RTHA patients had a Charlson Index ≥ 1 and this proportion rose to 24.6% in 2008 (p < 0.001). The ratio RTHA/THA remained stable and around 20% in Spain along the entire period

The crude overall in-hospital mortality (IHM) increased from 1.16% in 2001 to 1.77% (p = 0.025) in 2008. For both sexes the risk of death was higher with age, with the highest mortality rates found among those aged 85 or over. After multivariate analysis no change was observed in IHM over time. The mean inflation adjusted cost per patient increased by 78.3%, from 9, 375 to 16, 715 Euros from 2001 to 2008.

After controlling for possible confounders using Poisson regression models, we observed that the incidence of RTHA hospitalizations significantly increased for men and women over the period 2001 to 2008 (IRR 1.10, 95% CI 1.03-1.18 and 1.08, 95% CI 1.02-1.14 respectively).

Conclusions

The crude incidence of RTHA in Spain showed a small but significant increase from 2001 to 2008 with concomitant reductions in LOS, significant increase in co-morbidities and cost per patient.

Patient, physician, encounter, and billing characteristics predict the accuracy of syndromic surveillance case definitions

BACKGROUND

Syndromic surveillance systems are plagued by high false-positive rates. In chronic disease monitoring, investigators have identified several factors that predict the accuracy of case definitions based on diagnoses in administrative data, and some have even incorporated these predictors into novel case detection methods, resulting in a significant improvement in case definition accuracy. Based on findings from these studies, we sought to identify physician, patient, encounter, and billing characteristics associated with the positive predictive value (PPV) of case definitions for 5 syndromes (fever, gastrointestinal, neurological, rash, and respiratory (including influenza-like illness)).

METHODS

The study sample comprised 4,330 syndrome-positive visits from the claims of 1,098 randomly-selected physicians working in Quebec, Canada in 2005-2007. For each visit, physician-facilitated chart review was used to assess whether the same syndrome was present in the medical chart (gold standard). We used multivariate logistic regression analyses to estimate the association between claim-chart agreement about the presence of a syndrome and physician, patient, encounter, and billing characteristics.

RESULTS

The likelihood of the medical chart agreeing with the physician claim about the presence of a syndrome was higher when the treating physician had billed many visits for the same syndrome recently (ORper 10 visit, 1.05; 95% CI, 1.01-1.08), had a lower workload (ORper 10 claims, 0.93; 95% CI, 0.90-0.97), and when the patient was younger (ORper 5 years of age, 0.96; 95% CI, 0.94-0.97), and less socially deprived (ORmost versus least deprived, 0.76; 95% CI, 0.60-0.95).

CONCLUSIONS

Many physician, patient, encounter, and billing characteristics associated with the PPV of surveillance case definition are accessible to public health, and could be used to reduce false-positive alerts by surveillance systems, either by focusing on the data most likely to be accurate, or by adjusting the observed data for known biases in diagnosis reporting and performing surveillance using the adjusted values.

Use of UpToDate and outcomes in US hospitals

BACKGROUND

Computerized clinical knowledge mana-gement systems hold enormous potential for improving quality and efficiency. However, their impact on clinical practice is not well known.

OBJECTIVE

To examine the impact of UpToDate on outcomes of care.

DESIGN

Retrospective study.

SETTING

National sample of US inpatient hospitals.

PATIENTS

Fee-for-service Medicare beneficiaries.

INTERVENTION

Adoption of UpToDate in US hospitals.

MEASUREMENT

Risk-adjusted lengths of stay, mortality rates, and quality performance.

RESULTS

We found that patients admitted to hospitals using UpToDate had shorter lengths of stay than patients admitted to non-UpToDate hospitals overall (5.6 days vs 5.7 days; P < 0.001) and among 6 prespecified conditions (range, -0.1 to -0.3 days; P < 0.001 for each). Further, patients admitted to UpToDate hospitals had lower risk-adjusted mortality rate for 3 of the 6 conditions (range, -0.1% to -0.6% mortality reduction; P < 0.05). Finally, hospitals with UpToDate had better quality performance for every condition on the Hospital Quality Alliance metrics. In subgroup analyses, we found that it was the smaller hospitals and the non-teaching hospitals where the benefits of the UpToDate seemed most pronounced, compared to the larger, teaching institutions where the benefits of UpToDate seemed small or nonexistent.

CONCLUSIONS

We found a very small but consistent association between use of UpToDate and reduced length of stay, lower risk-adjusted mortality rates, and better quality performance, at least in the smaller, non-teaching institutions. These findings may suggest that computerized tools such as UpToDate could be helpful in improving care.

Statin use and musculoskeletal pain among adults with and without arthritis

BACKGROUND

Musculoskeletal symptoms are common adverse effects of statins, yet little is known about the prevalence of musculoskeletal pain and statin use in the general population.

METHODS

We conducted a cross-sectional study of the National Health and Nutrition Examination Survey 1999-2004. We estimated the prevalence of self-reported musculoskeletal pain according to statin use and calculated prevalence ratio estimates of musculoskeletal pain obtained from adjusted multiple logistic regression modeling.

RESULTS

Among 5170 participants without arthritis, the unadjusted prevalence of musculoskeletal pain was significantly higher for statin users reporting pain in any region (23% among statin users, 95% confidence interval [CI], 19-27, compared with 18% among those not using statins, 95% CI, 17-20; P=.02) and in the lower extremities (12% among statin users, 95% CI, 8-16, compared with 8% among those not using statins, 95% CI, 7-9; P=.02). Conversely, among 3058 participants with arthritis, statin use was not associated with higher musculoskeletal pain in any region. After controlling for confounders, among those without arthritis, statin use was associated with a significantly higher prevalence of musculoskeletal pain in any region, the lower back, and the lower extremities (adjusted prevalence ratios: 1.33 [CI, 1.06-1.67]; 1.47 [CI, 1.02-2.13]; 1.59 [CI, 1.12-2.22], respectively). Among participants with arthritis, no association was observed between musculoskeletal pain and statin use on adjusted analyses.

CONCLUSION

In this population-based study, statin use was associated with a higher prevalence of musculoskeletal pain, particularly in the lower extremities, among individuals without arthritis. Evidence that statin use was associated with musculoskeletal pain among those with arthritis was lacking.

Patient-related risk factors for postoperative mortality and periprosthetic joint infection in medicare patients undergoing TKA

BACKGROUND

The impact of specific baseline comorbid conditions on the relative risk of postoperative mortality and periprosthetic joint infection (PJI) in elderly patients undergoing TKA has not been well defined.

QUESTIONS/PURPOSES

We calculated the relative risk of postoperative mortality and PJI associated with 29 comorbid conditions in Medicare patients undergoing TKA.

PATIENTS AND METHODS

The Medicare 5% sample was used to calculate the relative risk of 90-day postoperative mortality and PJI as a function of 29 preexisting comorbid conditions in 83,011 patients who underwent primary TKA between 1998 and 2007.

RESULTS

The independent risk factors for 90-day postoperative mortality (in decreasing order of significance) were congestive heart failure, metastatic cancer, renal disease, peripheral vascular disease, cerebrovascular disease, lymphoma, cardiac arrhythmia, dementia, pulmonary circulation disorders, and chronic liver disease. The independent risk factors for PJI (in decreasing order of significance) were congestive heart failure, chronic pulmonary disease, preoperative anemia, diabetes, depression, renal disease, pulmonary circulation disorders, obesity, rheumatologic disease, psychoses, metastatic tumor, peripheral vascular disease, and valvular disease.

CONCLUSIONS

We believe this information important when counseling elderly patients regarding the risks of mortality and PJI after TKA and risk-adjusting publicly reported TKA patient outcomes.

Variations in the use of internal fixation for distal radial fracture in the United States medicare population

BACKGROUND

Distal radial fractures affect an estimated 80,000 elderly Americans each year. Although the use of internal fixation for the treatment of distal radial fractures is becoming increasingly common, there have been no population-based studies to explore the dissemination of this technique. The aims of our study were to determine the current use of internal fixation for the treatment of distal radial fractures in the Medicare population and to examine regional variations and other factors that influence use of this treatment. We hypothesized that internal fixation of distal radial fractures would be used less commonly in male and black populations compared with other populations because the prevalence of osteoporosis is lower in these populations, and that use of internal fixation would be correlated with the percentage of the patients who were treated by a hand surgeon in a particular region.

METHODS

We performed an analysis of complete 2007 Medicare data to determine the percentage of distal radial fractures that were treated with internal fixation in each hospital referral region. We then analyzed the association of patient and physician factors with the type of fracture treatment received, both nationally and within each hospital referral region.

RESULTS

We identified 85,924 Medicare beneficiaries with a closed distal radial fracture who met the inclusion criteria, and 17.0% of these patients were treated with internal fixation. Fractures were significantly less likely to be treated with internal fixation in men than in women (odds ratio, 0.84; 95% confidence interval, 0.80 to 0.89) and in black patients than in white patients (odds ratio, 0.74; 95% confidence interval, 0.65 to 0.85). Patients were more likely to be treated with internal fixation rather than with another treatment if they were treated by a hand surgeon than if they were treated by an orthopaedic surgeon who was not a hand surgeon (odds ratio, 2.49; 95% confidence interval, 2.29 to 2.70). Use of internal fixation ranged from 4.6% to 42.1% (nearly a ten-fold difference) among hospital referral regions. The percentage of patients treated with internal fixation within a hospital referral region was positively correlated with the percentage of patients in that region who were treated by a hand surgeon (correlation coefficient, 0.34; p &lt; 0.0001).

CONCLUSIONS

The use of internal fixation for the treatment of a distal radial fracture differs widely among geographical regions and patient populations. Such variations highlight the need for improved comparative-effectiveness data to guide the treatment of this fracture.

Identifying chronic conditions in Medicare claims data: evaluating the Chronic Condition Data Warehouse algorithm

OBJECTIVE

To examine the strengths and limitations of the Center for Medicare and Medicaid Services' Chronic Condition Data Warehouse (CCW) algorithm for identifying chronic conditions in older persons from Medicare beneficiary data.

DATA SOURCES

Records from participants of the NHANES I Epidemiologic Follow-up Study (NHEFS 1971-1992) linked to Medicare claims data from 1991 to 2000.

STUDY DESIGN

We estimated the percent of preexisting cases of chronic conditions correctly identified by the CCW algorithm during its reference period and the number of years of claims data necessary to find a preexisting condition.

PRINCIPAL FINDINGS

The CCW algorithm identified 69 percent of preexisting diabetes cases but only 17 percent of preexisting arthritis cases. Cases identified by the CCW are a mix of preexisting and newly diagnosed conditions.

CONCLUSIONS

The prevalence of conditions needing less frequent health care utilization (e.g., arthritis) may be underestimated by the CCW algorithm. The CCW reference periods may not be sufficient for all analytic purposes.

Derivation and preliminary validation of an administrative claims-based algorithm for the effectiveness of medications for rheumatoid arthritis

INTRODUCTION

Administrative claims data have not commonly been used to study the clinical effectiveness of medications for rheumatoid arthritis (RA) because of the lack of a validated algorithm for this outcome. We created and tested a claims-based algorithm to serve as a proxy for the clinical effectiveness of RA medications.

METHODS

We linked Veterans Health Administration (VHA) medical and pharmacy claims for RA patients participating in the longitudinal Department of Veterans Affairs (VA) RA registry (VARA). Among individuals for whom treatment with a new biologic agent or nonbiologic disease-modifying agent in rheumatic disease (DMARD) was being initiated and with registry follow-up at 1 year, VARA and administrative data were used to create a gold standard for the claims-based effectiveness algorithm. The gold standard outcome was low disease activity (LDA) (Disease Activity Score using 28 joint counts (DAS28) ≤ 3.2) or improvement in DAS28 by > 1.2 units at 12 ± 2 months, with high adherence to therapy. The claims-based effectiveness algorithm incorporated biologic dose escalation or switching, addition of new disease-modifying agents, increase in oral glucocorticoid use and dose as well as parenteral glucocorticoid injections.

RESULTS

Among 1,397 patients, we identified 305 eligible biologic or DMARD treatment episodes in 269 unique individuals. The patients were primarily men (94%) with a mean (± SD) age of 62 ± 10 years. At 1 year, 27% of treatment episodes achieved the effectiveness gold standard. The performance characteristics of the effectiveness algorithm were as follows: positive predictive value, 76% (95% confidence interval (95% CI) = 71% to 81%); negative predictive value, 90% (95% CI = 88% to 92%); sensitivity, 72% (95% CI = 67% to 77%); and specificity, 91% (95% CI = 89% to 93%).

CONCLUSIONS

Administrative claims data may be useful in evaluating the effectiveness of medications for RA. Further validation of this effectiveness algorithm will be useful in assessing its generalizability and performance in other populations.

The burden of depression in prostate cancer

OBJECTIVE

We sought to analyze the prevalence and incremental burden of depression among elderly with prostate cancer.

METHODS

We adopted a retrospective cohort design using the Surveillance, Epidemiology and End Results-Medicare linked database between 1995 and 2003. Patients with prostate cancer diagnosed between 1995 and 1998 were identified and followed retrospectively for 1 year pre-diagnosis and up to 8 years post diagnosis. In this cohort of patients with prostate cancer, depression during treatment phase (1 year after diagnosis of prostate cancer) or in the follow-up phase was identified using the International Classification of Diseases-Ninth Revision depression-related codes. Poisson, general linear (log-link) and Cox regression models were used to determine the association between depression status during treatment and follow-up phases and outcomes-health resource utilization, cost and mortality.

RESULTS

Of the 50,147 patients newly diagnosed with prostate cancer, 4285 (8.54%) had a diagnosis of depression. A diagnosis of depression during treatment phase was associated with higher odds of emergency room visits (odds ratio (OR) = 4.45, 95% CI = 4.13, 4.80), hospitalizations (OR = 3.22, CI = 3.08, 3.37), outpatient visits (OR = 1.71, CI = 1.67, 1.75) and excess risk of death over the course of the follow-up interval (hazard ratio = 2.82, CI = 2.60, 3.06). Health care costs associated with depression remained elevated compared with costs for men without depression, over the course of the follow-up.

CONCLUSIONS

Depression during the treatment phase was associated with significant health resource utilization, costs and mortality among men with prostate cancer. These findings emphasize the need to effectively identify and treat depression in the setting of prostate cancer.

The relationship between ASSH membership and the treatment of distal radius fracture in the United States Medicare population

PURPOSE

Internal fixation for distal radius fractures (DRFs) in the elderly has increased from 3% in 1997 to 17% in 2007. This increase has been uneven across regions of the United States. There is some evidence that patients treated by hand surgeons receive internal fixation at an increased rate and that hand surgeons might be driving the increased usage in regions where their presence is greatest. The specific aim of this study was to explore this relationship by analyzing Medicare beneficiaries treated by members of the American Society for Surgery of the Hand (ASSH).

METHODS

Surgeons who were members of ASSH in 2007 were matched with surgeons treating Medicare beneficiaries for DRFs in the same year. We then fit a series of multilevel models to estimate the proportion of total variance in internal fixation usage explained by ASSH membership status, patient demographic data, patient comorbidity, and/or type of fracture diagnosed.

RESULTS

Beneficiaries treated by ASSH members received internal fixation significantly more often than beneficiaries who were treated by surgeons who were not ASSH members. ASSH member status accounts for 12% of the total variance in internal fixation utilization.

CONCLUSIONS

Medicare beneficiaries who were treated by ASSH member surgeons receive internal fixation at a significantly higher rate than do patients of other physicians. When there is uncertainty about the optimal treatment for a condition, there is the possibility for specialty-related disparities. This specialty effect contributes to the national variations in the treatment of DRFs in the Medicare population.