Algorithms can be used to identify fragility fracture cases in physician-claims databases

Jean, Sonia; Candas, Bernard; Belzile, Étienne L.; Morin, Suzanne N; Bessette, Louis; Dodin, Sylvie; Brown, Jacques P.

doi:10.1007/s00198-011-1559-4

Cited by 59 publications

(78 citation statements)

References 33 publications

(59 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Fracture ascertainment from administrative data sources may be incomplete, particularly for vertebral fractures, although similar algorithms have proven useful for both vertebral and nonvertebral fracture identification. (37,38) We used a FRAX tool that has been directly validated in the Canadian population, (26,27) but there was incomplete information on some of the baseline clinical risk factors (eg, parental hip fracture), and for others proxies were used (eg, smoking, alcohol intake) as previously described. (25) Despite these limitations, predicted 10-year fracture probability agreed very closely with the observed fracture incidence estimated to 10 years among untreated women, suggesting reasonably complete ascertainment of fractures and risk factors.…”

Section: Discussionmentioning

confidence: 99%

Does osteoporosis therapy invalidate FRAX for fracture prediction?

Leslie

Lix

Johansson

et al. 2012

Journal of Bone and Mineral Research

112

View full text Add to dashboard Cite

Ten-year fracture risk assessment with the fracture risk assessment system (FRAX) is increasingly used to guide treatment decisions. Osteoporosis pharmacotherapy reduces fracture risk, but the effect is greater than can be explained from the increase in bone mineral density (BMD). Whether this invalidates fracture predictions with FRAX is uncertain. A total of 35,764 women (age !50 years) and baseline BMD testing (1996)(1997)(1998)(1999)(2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007) had FRAX probabilities retroactively calculated. A provincial pharmacy database was used to identify osteoporosis medication use. Women were categorized as untreated, current high adherence users [medication possession ratio (MPR) !0.80 in the year after BMD testing], current low adherence users (MPR <0.80), and past users. Fractures outcomes to 10 years were established form a population-based health data repository. FRAX and femoral neck BMD alone stratified major osteoporotic and hip fracture risk within untreated and each treated subgroup (all p-values <0.001) with similar area under the receiver operating characteristic curve. In untreated and each treated subgroup, a stepwise gradient in observed 10-year major osteoporotic and hip fracture incidence was found as a function of the predicted probability tertile (all p-values <0.001 for linear trend). Concordance (calibration) plots for major osteoporotic fractures and hip fractures showed good agreement between the predicted and observed 10-year fracture incidence in untreated women and each treated subgroup. Only in the highest risk tertile of women highly adherent to at least 5 years of bisphosphonate use was observed hip fracture risk significantly less than predicted, though major osteoporotic fracture risk was similar to predicted. In summary, this work suggests that the FRAX tool can be used to predict fracture probability in women currently or previously treated for osteoporosis. Although FRAX should not be used to assess the reduction in fracture risk in individuals on treatment, it may still have value for guiding the need for continued treatment or treatment withdrawal. ß

show abstract

Section: Discussionmentioning

confidence: 99%

Does osteoporosis therapy invalidate FRAX for fracture prediction?

Leslie

Lix

Johansson

et al. 2012

Journal of Bone and Mineral Research

112

View full text Add to dashboard Cite

show abstract

“…(19) This cohort of recently-fractured women was identified from a previous study that validated an algorithm for fracture case identification. (20) Administrative databases of healthcare services (Ré gie de l'assurance maladie du Qué bec [RAMQ] physicians' fee-for-service claims and MedEcho hospital discharge) delivered to the majority of residents in Quebec were considered to track fracture-related direct medical resource utilization in the year after fracture. RAMQ databases collect and store medical claims information from approximately 94% of Quebec-based physicians and on almost 98% of 7.5 million residents of the province.…”

Section: Study Cohort and Designmentioning

confidence: 99%

“…(24) Identification of incident fracture cases An algorithm, developed using the physician fee-for-service claims database (RAMQ), identified all incident nonvertebral fracture cases during the period of interest. (20) The algorithm is designed to first select all medical services claims potentially associated with fracture treatment: (1) claims with medical services billing codes definitively related to fracture care (i.e., open or closed reduction), or (2) claims with medical service billing codes not limited to fracture care (ie, immobilization, consultation, principal or follow-up visit with an orthopedic surgeon [OS], emergency physician [EP], or general practitioner [GP]) if they were combined with ICD-9-CM diagnostic codes of fracture. The algorithm considered an incident fracture had occurred if there was at least one claim associated with (1) fracture treatment (open reduction, closed reduction, immobilization), (2) principal visit to an OS with at least one other claim, or (3) consultation with an OS with at least one other claim.…”

Section: Data Sourcesmentioning

confidence: 99%

“…The validity of this algorithm has been evaluated and published. (20) Fracture sites were defined by the specific medical service codes of the index claim related to the treatment of fracture or, if not specific to the treatment of fracture, to the ICD-9-CM codes. To establish the complete temporal sequence of medical care for each fracture, the algorithm identified any other claim (emergency room visit, follow-up visit, etc.)…”

Section: Data Sourcesmentioning

confidence: 99%

“…Vertebra, sacrum, and coccyx fractures were not included in the analyses because sensitivity (Sn) of the algorithm was unacceptably low for the identification of these fractures. (20) Furthermore, vertebral fractures are generally underreported in administrative databases. (27) Rib and sternum fractures were excluded because there are very few specific medical service codes (open and closed reduction) and no immobilization related to the treatment of these fractures and codes related to surgeries at these sites may be used to treat patients after surgical intervention related to cardiovascular problems or likely define traumatic fracture rather than fragility fracture.…”

Section: Data Sourcesmentioning

confidence: 99%

See 2 more Smart Citations

Direct medical resource utilization associated with osteoporosis-related nonvertebral fractures in postmenopausal women

Jean¹,

Bessette²,

Belzile³

et al. 2012

Journal of Bone and Mineral Research

View full text Add to dashboard Cite

The purposes of this study were to assess direct medical resource utilization related to the treatment of nonvertebral osteoporotic fractures within 1 year postfracture and to evaluate whether age impacts resource utilization. A previously-validated algorithm for physician claims databases identified 15,327 women aged 50 years or older with incident fracture at nonvertebral osteoporotic sites between January 1, 2004 and December 31, 2005. Administrative databases of the health services available to all residents in Quebec served to study fracture-related health resource utilization in the year after fracture. Data were linked by a unique personal identifier, creating a longitudinal cohort of all fracture cases for health resource utilization. The proportions of fractures treated by open reduction, closed reduction, immobilization or follow-up by an orthopedic surgeon (OS) were evaluated. The mean number of claims for consultation with an OS or other clinicians in inpatient and outpatient visits, the hospitalization rate and length of stay (LOS) were assessed. Hip/femur fractures represented the highest rate of resource utilization because the majority of them required surgery (91.1%) and hospitalization (94.5%) with a mean (median) LOS of 39.2 (31) days. However, other nonvertebral fracture types needed significant clinical care related to surgery (27.9%), follow-up consultation with an OS (77.6%), and hospitalization (27.3% of total LOS). Even pelvic fractures, which often do not require surgical treatment, commanded high resource utilization due to the high hospitalization rate (67.4%) with mean (median) LOS of 34.2 (26) days. Moreover, age was an important determinant of health resource utilization, being associated with an increased number of visits to other physicians, hospitalization, and length of hospitalization (LOS), admissions to long term care (LTC), and death. Osteoporosis-related fractures accounted for substantial healthcare resource utilization. With an aging population and increased prevalence of fractures, strategies for osteoporosis management need to be introduced to reduce the healthcare burden. ß

show abstract

Incidence Rate and Factors Associated With Fractures Among Medicare Beneficiaries With Ankylosing Spondylitis in the United States

Stovall,

Kersey,

et al. 2023

Arthritis Care & Research

View full text Add to dashboard Cite

ObjectiveWe evaluated the incidence rate and factors associated with fractures among adults with ankylosing spondylitis (AS).MethodsWe performed a retrospective cohort study with data from the Rheumatology Informatics System for Effectiveness registry linked to Medicare claims from 2016 to 2018. Patients were required to have two AS International Classification of Diseases codes 30 or more days apart and a subsequent Medicare claim. Then, 1 year of baseline characteristics were included, after which patients were observed for fractures. First, we calculated the incidence rate of fractures. Second, we constructed logistic regression models to identify factors associated with the fracture, including age, sex, race and ethnicity, body mass index, Medicare/Medicaid dual eligibility, area deprivation index, Charlson comorbidity index, smoking status, osteoporosis, historical fracture, and use of osteoporosis treatment, glucocorticoids, and opioids.ResultsWe identified 1,426 adults with prevalent AS. Mean ± SD age was 69.4 ± 9.8 years, 44.3% were female, and 77.3% were non‐Hispanic White. Fractures occurred in 197 adults with AS. The overall incidence rate of fractures was 76.7 (95% confidence interval [CI] 66.4–88.6) per 1,000 person‐years. Older age (odds ratio [OR] 2.8, 95% CI 1.39–5.65), historical fracture (OR 5.24, 95% CI 3.44–7.99), and use of more than 30 mg morphine equivalent (OR 1.86, 95% CI 1.08–3.19) conferred increased odds of fracture.ConclusionsIn this large sample of Medicare beneficiaries with AS, increasing age, historical fracture, and use of opioids had higher odds of fracture. Men and women were equally likely to have a fracture. Because opioid use was associated with fracture in AS, this high‐risk population should be considered for interventions to mitigate risk.

show abstract

Algorithms can be used to identify fragility fracture cases in physician-claims databases

Cited by 59 publications

References 33 publications

Does osteoporosis therapy invalidate FRAX for fracture prediction?

Does osteoporosis therapy invalidate FRAX for fracture prediction?

Direct medical resource utilization associated with osteoporosis-related nonvertebral fractures in postmenopausal women

Incidence Rate and Factors Associated With Fractures Among Medicare Beneficiaries With Ankylosing Spondylitis in the United States

Contact Info

Product

Resources

About