RESEARCH - Use of VA pharmacoepidemiology database to define the scope of the problem of steroid-induced fractures

[Guest guest]

Massachusetts Veterans Epidemiology Research and Information Center

" Use of VA pharmacoepidemiology database to define the scope of the problem

of steroid-induced fractures " :

BACKGROUND: The increased risk for fractures associated with chronic steroid

use results in increased morbidity and healthcare utilization. We selected

this common problem to evaluate the benefits of developing a

pharmacoepidemiology database using the Veterans Integrated Service Network

of New England (VISN 1). This study describes the methods for developing

this database and the scope of the problem of steroid induced fractures.

METHODS: The unvalidated database was established using VISN 1 inpatient,

outpatient administrative treatment and pharmacy files between January 1998

and November 2001. Prednisone comprised >80% of the VA steroid prescriptions

and was used to define steroid users. All other steroid users were included

with the non-steroid users. Chronic and occasional prednisone users were

defined as individuals prescribed prednisone for >= 3 months or <3 months

during any 12 month period, respectively. Individuals were identified by

ICD-codes for fractures occurring after prednisone use and for the

occurrence of a fracture and a diagnosis of osteoporosis. Differences

between groups were determined by Chi-square analysis and are reported as

relative risks (RR) with 95% confidence intervals (CI). Mantel-Haenszel

chi-square test for trend was used to the determined the presence of a

linear trend in risk for fracture in the three groups (chronic versus

occasional prednisone use versus no prednisone use).

RESULTS: 12,780 individuals out of 203,170 VA pharmacy users were identified

as receiving a prescription for prednisone (37.9% chronic user, 62.1%

occasional user). The occurrence of fractures in the prednisone group

compared to all other VA pharmacy users was 3.48% versus 2.98% (RR 1.17; CI

1.06-1.20). There were more fractures among chronic prednisone users

compared to non-users(RR 4.43 95% CI 3.40-5.83). There was a significant

trend in the risk for fractures among chronic and occasional prednisone

users versus non-users (p <0.001). The diagnosis of both fracture and

osteoporosis was also common among prednisone users and even greater among

chronic users (RR 4.43 95% CI 3.40-5.83 and RR 7.63 95% CI 5.60-10.41,

respectively). The test for trend comparing non-prednisone users and chronic

or occasional prednisone users was again significant (p<0.001).

CONCLUSION: The common occurrence of fractures among individuals using

prednisone is similar to other published reports. Establishment of a VA

pharmacoepidemiology database with well-defined methods enabled us to detail

the magnitude of risk in our patient population. Use of such a database will

help define problem areas and allow for the assessment of the

appropriateness and successfulness of preventive interventions.

http://www.maveric.org/route/maveric/pharmaco/csteroid.asp

Not an MD

I'll tell you where to go!

Mayo Clinic in Rochester

http://www.mayoclinic.org/rochester

s Hopkins Medicine

http://www.hopkinsmedicine.org