SEARCH

SEARCH BY CITATION

Keywords:

  • Osteoporosis;
  • Rheumatoid arthritis;
  • Quality of care;
  • Glucocorticoid-induced osteoporosis

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. REFERENCES

Objective

Osteoporosis in patients with rheumatoid arthritis (RA) is increasingly recognized as a major comorbidity. We examined past management patterns for glucocorticoid-induced osteoporosis and attempted to improve care through an educational intervention. The goal was to examine the frequency of osteoporosis management in patients with RA treated at a large academic rheumatology practice.

Methods

We performed a structured chart review on randomly selected patients seen during 2004 for RA. Osteoporosis management was defined as a bone mineral density test or receipt of a medication for osteoporosis in the prior 24 months. The frequency of osteoporosis management among our study group was assessed. We also examined how glucocorticoid dosage affected osteoporosis management in adjusted models.

Results

We reviewed the records for 193 patients, 99 had not used glucocorticoids in the prior 24 months, and 94 had used them. Of the total study population, 48% had received a bone mineral density test or medication for osteoporosis. Some form of osteoporosis management was present for 64% of patients taking ≥5 mg prednisone for ≥3 months compared with 38% for patients taking none (P = 0.002). Predictors of osteoporosis management included older age, female sex, glucocorticoid dosage, and prior osteoporosis diagnosis or fracture.

Conclusion

The frequency of osteoporosis management appears to have increased compared with a prior chart review.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. REFERENCES

Substantial data support the association between osteoporosis and glucocorticoids. Glucocorticoids appear to both reduce bone mineral density (BMD) and increase the risk of fractures (1, 2). The risk of osteoporosis appears associated with even very low dose glucocorticoids, i.e., <2.5 mg/day of prednisolone (3). A committee of the American College of Rheumatology (ACR) recognized the risk of glucocorticoid use and recommended screening for osteoporosis with BMD testing, and subsequent treatment in patients with T scores less than −1.0 (4). Glucocorticoid-induced osteoporosis is especially worrisome in patients with rheumatoid arthritis (RA) because of the already increased risk of diminished BMD and fractures observed in patients with RA (5). In addition, the use of glucocorticoids may increase in light of recently published evidence that low dose glucocorticoid therapy is an effective disease-modifying agent in the management of RA (6, 7).

Several studies have demonstrated that management of the risk of osteoporosis in patients taking glucocorticoids is suboptimal (8, 9). Although patients seen by rheumatologists are more likely to receive screening and/or treatment (10), we recently documented suboptimal management of glucocorticoid-induced osteoporosis in patients with RA in our academic rheumatology practice. Specifically, 23% of patients taking at least 5 mg of prednisone for ≥3 months underwent BMD testing, 45% of patients received a prescription medication for osteoporosis, and 51% of the patients had either undergone BMD testing or received a prescription medication (11). These results led us to develop and test an educational intervention for rheumatologists in a randomized controlled trial; the trial failed to demonstrate efficacy for the intervention (12).

Prior to further quality improvement efforts, we performed the current chart review to determine the frequency of osteoporosis management in patients with RA and to determine whether there are subgroups of patients who are more or less likely to undergo screening and/or treatment. We included patients with RA who were and were not taking glucocorticoids to understand whether the use of this medication affected the frequency of osteoporosis management. Also, we focused on patients with RA taking ≥5 mg of prednisone for ≥3 months (the ACR-recommended threshold for screening and/or treatment [4]) to determine whether rheumatologists in our practice were adhering to ACR recommendations. Answers to these questions would help us decide whether to pursue a new quality improvement intervention and which patients to target.

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. REFERENCES

Role of the study sponsor.

Procter & Gamble had no role in the study design, data collection, data analysis, or writing of the manuscript.

Patient population.

We searched for all patients seen in our rheumatology practice in 2004 who had a diagnosis code consistent with RA (International Classification of Diseases, Ninth Revision 714.xx). From this cohort, we randomly selected patients whose rheumatologist had documented a diagnosis of RA in the medical record and who had at least 2 visits for RA in 2004. The last visit in 2004 was defined as the index date. Patients seen by 1 of the 4 rheumatologists involved in this study (DHS, JNK, JFB, JSC) were not included. Because we were interested in osteoporosis management in patients with RA treated with and without glucocorticoids, patients were not required to have used glucocorticoids to be included in the study cohort. The Partners Healthcare Institutional Review Board approved the study protocol.

Data collection.

A structured data abstraction form (available upon request) was used to review medical records for the 24 months prior to the index date. The abstraction form included information on patient characteristics (age, sex), comorbid illnesses (hypertension, diabetes, ischemic heart disease, cancer, depression, liver disease [hepatitis, cirrhosis], chronic kidney disease/insufficiency, asthma/chronic obstructive pulmonary disease, stroke/transient ischemic attack, chronic alcohol abuse, thyroid disease, and prior organ transplant), RA disease variables (duration of disease, serologic status, and treatments), osteoporosis treatments, BMD testing and results, and glucocorticoid dosage. Glucocorticoid dosages were converted into prednisone equivalents. Osteoporosis treatments included prescription medications such as alendronate, calcitonin, hormone therapy, raloxifene, risedronate, and teriparatide. All rheumatologist notes for the 24 months prior to the index date were searched for these items. The radiology reports were also searched for the 24 months. BMD testing that was documented in the radiology reports or in the physician's notes was also included.

Statistical analyses.

The characteristics of patients in the study cohort were examined. The primary outcome of interest for these analyses was receiving osteoporosis management, defined as undergoing a BMD test and/or receipt of a medication for osteoporosis; BMD testing and receipt of a medication were considered separately in secondary analyses. We assessed associations between these outcomes and key patient characteristics, including sex, age, duration of RA, use of a disease-modifying antirheumatic drug (DMARD), comorbid conditions, and glucocorticoid dosage. We created 3 ordered categories of glucocorticoid dosage: ≥5 mg of prednisone for ≥3 consecutive months, a lower dosage or shorter duration of oral glucocorticoids, and no glucocorticoid use. These associations were examined using chi-square test, with the exception of glucocorticoid use that was tested using Cochran-Armitage one-sided trend test. Finally, we examined patient factors as potential predictors of osteoporosis management using logistic regression. The unadjusted relationship between each variable and the primary outcomes was assessed in univariate logistic regression. Since we hypothesized possible associations between all variables and the outcomes, all variables were placed in a single multivariate model. The analyses accounted for clustering of patients within a given doctor's practice by using generalized estimating equations. Statistical analysis was performed with SAS, version 9.0 (SAS Institute, Cary, NC).

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. REFERENCES

A total of 233 patient charts were reviewed. Forty patients did not meet the inclusion criteria (16 had juvenile onset arthritis, 9 had other diagnoses, and 15 did not have 2 physician visits for RA). Our final sample included 193 patients with RA; 99 (51%) had taken oral glucocorticoids at some point during the 2-year study period and 94 had not taken oral glucocorticoids. The mean age of the subjects was 60 years, and 79% were female. Of the patients tested, 74% were positive for rheumatoid factor or anti-cyclic citrullinated peptide antibody. DMARDs were currently or previously taken by 92% of the patients, and 15% of the patients had either osteoporosis or a prior fracture.

The frequency of osteoporosis management is shown in Table 1. Of the total study population, 92 (48%; 95% confidence interval [95% CI] 41–56) received either a BMD test or a medication for osteoporosis in the 24 months prior to the index date. Osteoporosis management (defined as a BMD test or receiving medications in the prior 24 months) was more common in women than in men (P < 0.001), with women older than 50 years of age receiving management more frequently than those younger (P < 0.001). In addition, patients 50–64 years of age were more likely to have received osteoporosis management than those who were either younger or older (P < 0.001). Patients with longer RA duration and those with more comorbid conditions had a tendency toward more frequent osteoporosis management. There was a statistically significant trend toward more frequent osteoporosis management among patients taking more oral glucocorticoids (38% in those not taking oral glucocorticoids and 64% in those taking ≥5 mg of prednisone for ≥3 months; P = 0.002). Patients with a prior fracture or diagnosis of osteoporosis were much more likely to have received management (P < 0.001), but DMARD use was not a significant predictor of osteoporosis management.

Table 1. Bone mineral density (BMD) testing and osteoporosis medication use in patients with rheumatoid arthritis (RA)*
 TotalBMDMedicationBoth BMD and medicationEither BMD or medicationP
  • *

    Values are the number (percentage). P values from chi-square test or from Cochran-Armitage one-sided trend test for glucocorticoid use. DMARD = disease-modifying antirheumatic drug.

  • Hypertension, diabetes, ischemic heart disease, cancer, depression, liver disease (hepatitis, cirrhosis), chronic kidney disease/insufficiency, asthma/chronic obstructive pulmonary disease, stroke/transient ischemic attack, chronic alcohol abuse, thyroid disease, and prior organ transplant.

  • Includes all glucocorticoid use of <5 mg or <3 months.

All patients19370 (36)67 (35)45 (23)92 (48) 
Sex     < 0.001
 Male404 (10)6 (15)3 (8)7 (18) 
 Female15366 (43)61 (40)42 (27)85 (56) 
  <50 years4810 (21)6 (13)3 (6)13 (27)< 0.001
  ≥50 years10556 (53)55 (52)39 (37)72 (69) 
RA duration, years     0.14
 <58422 (26)23 (27)10 (12)35 (42) 
 ≥510948 (44)44 (40)35 (32)57 (52) 
Comorbid illness     0.16
 None7519 (25)23 (31)11 (15)31 (41) 
 ≥111851 (43)44 (37)34 (29)61 (52) 
Glucocorticoid use     0.002
 None9928 (28)32 (32)22 (22)38 (38) 
 Low dose4919 (39)17 (35)11 (22)25 (51) 
 ≥5 mg ≥3 months4523 (51)18 (40)12 (27)29 (64) 
Age, years     < 0.001
 <505310 (19)6 (11)3 (6)13 (25) 
 50–647135 (49)35 (49)26 (37)44 (62) 
 ≥656925 (36)26 (38)16 (23)35 (51) 
Prior osteoporosis or fracture     < 0.001
 Yes2919 (66)21 (72)16 (55)24 (83) 
 No16451 (31)46 (28)29 (18)68 (41) 
DMARD use     0.44
 Yes17764 (36)63 (36)41 (23)86 (49) 
 No166 (38)4 (25)4 (25)6 (38) 

The multivariable model showed that the following variables were all associated with an increased probability of osteoporosis management: age 50–64 years (adjusted odds ratio [OR] 6.9, 95% CI 2.2–21.7), age ≥65 (OR 4.5, 95% CI 1.5–13.6), female sex (OR 8.5, 95% CI 3.4–21.2), prior fracture or diagnosis of osteoporosis (OR 4.1, 95% CI 1.4–12.2), and use of ≥5 mg of prednisone for ≥3 months (OR 3.6, 95% CI 1.5–8.9) (Table 2).

Table 2. Adjusted predictors of osteoporosis management*
 Univariate ORMultivariate OR
  • *

    Values are OR (95% confidence interval). Multivariate OR calculated using a generalized estimating equation accounting for the clustering of patients. OR = odds ratio (reference category 1.00).

  • Hypertension, diabetes, ischemic heart disease, cancer, depression, liver disease (hepatitis, cirrhosis), chronic kidney disease/insufficiency, asthma/chronic obstructive pulmonary disease, stroke/transient ischemic attack, chronic alcohol abuse, thyroid disease, and prior organ transplant.

  • Both synthetic or biologic disease-modifying antirheumatic drugs (DMARD).

Age, years  
 <501.001.00
 50–645.0 (2.3–11.1)6.9 (2.2–21.7)
 ≥653.2 (1.4–6.9)4.5 (1.5–13.6)
Sex  
 Male1.001.00
 Female5.9 (2.5–14.1)8.5 (3.4–21.2)
Glucocorticoid use  
 None or <5 mg or <3 months1.001.00
 ≥5 mg ≥3 months2.4 (1.2–4.9)3.6 (1.5–8.9)
Prior osteoporosis or fracture6.8 (2.5–18.6)4.1 (1.4–12.2)
RA duration, years  
 <51.001.00
 ≥51.5 (0.9–2.7)1.2 (0.5–2.8)
Any comorbid illness1.5 (0.8–2.7)1.1 (0.6–2.3)
Any DMARD use1.6 (0.5–4.5)1.2 (0.4–3.6)

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. REFERENCES

Patients with RA are at high risk for osteoporosis, particularly those patients with RA that take glucocorticoids (5, 13). Although rheumatologists appear to be more attentive to screening and treating for glucocorticoid-induced osteoporosis than other physicians, in many settings it has been demonstrated that glucocorticoid-induced osteoporosis is suboptimally managed (8, 9). In our previous study of a large academic rheumatology practice using data from 1999, we found that 51% of 150 patients with RA taking ≥5 mg of prednisone for ≥3 months had undergone a BMD test or received a medication for osteoporosis (11). A randomized controlled trial of physician education and reminders conducted in this practice were not effective in increasing either screening or treatment (12). These past efforts motivated planning for a new intervention, including this study of osteoporosis management in 2004 in patients with RA seen in our practice. We anticipated that this chart review would allow us to understand whether there was suboptimal management of osteoporosis in our practice, and if specific patient groups were particularly poorly managed. We included patients taking and not taking oral glucocorticoids because of the growing body of research linking RA with osteoporosis.

Among those patients taking ≥5 mg of prednisone for ≥3 months, we found that the frequency of osteoporosis management was 64%. Comparing results from the cohorts assembled in 1999 (11) and 2004 (present study) is not simple because the cohorts were chosen using slightly different criteria (patients were required to only have 1 visit in the 1999 cohort, and 2 visits in the 2004 cohort). However, less than 5% of patients in the 1999 cohort had only 1 visit. The 13% increase (from 51% to 64%) in management between the 2 studies is not statistically significant (P = 0.12, chi-square test with 1 degree of freedom), but the power to detect a difference was limited by the relatively small number of patients in the current study taking ≥5 mg of prednisone for ≥3 months.

Although not statistically significant, the apparent 13% improvement may be clinically meaningful and merits discussion. The finding may have several explanations. First, increased physician and/or patient awareness may have contributed to the improvement. The ACR has published recommendations for glucocorticoid-induced osteoporosis since our last study (4). There have also been several trials published demonstrating the efficacy of bisphosphonates for glucocorticoid-induced osteoporosis (14, 15). In addition, our prior failed educational intervention may have increased awareness regarding this issue (12). Second, our patient populations differed slightly in age, sex, and seropositivity, and these differences may account for the apparent increase in management of osteoporosis. Third, several new rheumatologists have joined the practice since the first study. Although these new rheumatologists account for relatively few patients (n = 11), differences in their osteoporosis management may have contributed to the increase.

In addition to our inability to precisely compare results from our prior study with this one, our study has several other limitations. We focused on process measures (osteoporosis care) and not outcomes, such as BMD or fractures. A much larger study with longer followup would be required to appreciate differences in these outcome measures. These data come from only 1 practice and therefore may not reflect a larger trend. We encourage others to examine other rheumatology practices. Relying on medical records may have underestimated the orders for BMD tests and/or medications that went unfilled or were not recorded in the medical records available to us.

In our large academic rheumatology practice, we found that almost two-thirds of patients with RA taking ≥5 mg of prednisone for ≥3 months had evidence of osteoporosis management during a 2-year study period. This appears to be a modest improvement over results from 5 years ago. It is interesting to note that during the current 2-year study period less than 40% of patients with RA not taking glucocorticoids had undergone a BMD test or received a medication for osteoporosis. Several patient factors were associated with the likelihood of osteoporosis management including older age, female sex, a prior fracture or diagnosis of osteoporosis, and more intense use of oral glucocorticoids. All of these variables have been associated with an increased risk of fracture based on prior studies (16). We have focused our discussion on the group of patients taking at least 5 mg of prednisone for at least 3 months because there are clear management recommendations for this group. While almost one-third of such patients had no clear evidence for a BMD test or use of a medication for osteoporosis, concerns regarding the use of bisphosphonates and raloxifene in women of child-bearing potential may limit the widespread acceptance of screening and treatment recommendations. Moreover, we are not sure whether typical interventions (i.e., educational) would be successful at further increasing the management of osteoporosis among this group. Efforts to optimize medical care are rarely able to increase management above approximately 75%. Some options that might be more potent include point-of-care reminders and nurse-run services that operate based on simple algorithms. We look forward to continued efforts to measure and improve the quality of care in rheumatology practice.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. REFERENCES