Population-based assessment of adverse events associated with long-term glucocorticoid use

Arthritis & Rheumatism (Arthritis Care & Research)Vol. 55, No. 3, June 15, 2006, pp 420 – 426DOI 10.1002/art.21984 2006, American College of Rheumatology Population-Based Assessment of Adverse Events
Associated With Long-Term Glucocorticoid Use
JEFFREY R. CURTIS,1 ANDREW O. WESTFALL,1 JEROAN ALLISON,1 JOHANNES W. BIJLSMA,2
ALLISON FREEMAN,3 VARGHESE GEORGE,1 STACEY H. KOVAC,4 CLAIRE M. SPETTELL,3 AND
KENNETH G. SAAG1
Objective. The frequency of many adverse events (AEs) associated with low-dose glucocorticoid use is unclear. We
sought to determine the prevalence of glucocorticoid-associated AEs in a large US managed care population.
Methods. Using linked administrative and pharmacy claims, adults receiving >60 days of glucocorticoids were identi-
fied. These individuals were surveyed about glucocorticoid use and symptoms of 8 AEs commonly attributed to
glucocorticoid use.
Results. Of the 6,517 eligible glucocorticoid users identified, 2,446 (38%) returned the mailed survey. Respondents were
29% men with a mean ؎ SD age of 53 ؎ 14 years; 79% were white and 13% were African American. Respondents had
a mean ؎ SD of 7 ؎ 3 comorbid conditions and were prescribed a mean ؎ SD prednisone-equivalent dosage of 16 ؎ 14
mg/day. More than 90% of individuals reported at least 1 AE associated with glucocorticoid use; 55% reported that at least
1 AE was very bothersome. Weight gain was the most common self-reported AE (70% of the individuals), cataracts (15%)
and fractures (12%) were among the most serious. After multivariable adjustment, all AEs demonstrated a strong
dose-dependent association with cumulative glucocorticoid use. Among users of low-dose therapy (<7.5 mg of prednisone
per day), increasing duration of use was significantly associated with acne, skin bruising, weight gain, and cataracts.
Conclusion. The prevalence of 8 commonly attributed self-reported glucocorticoid-associated AEs was significantly
associated with cumulative and average glucocorticoid dose in a dose-dependent fashion. Physicians should be vigilant
for glucocorticoid-related AEs and should counsel patients about possible risks, even among low-dose long-term users.
KEY WORDS. Glucocorticoids; Adverse events; Cataracts; Fractures.
INTRODUCTION
emerging evidence supporting a disease-modifying role inrheumatoid arthritis (3– 6), concern for adverse events Glucocorticoids are estimated to be used long-term by (AEs) associated with glucocorticoids often limits their 0.5–1% of the general population and up to 2.5% of older use. For many individuals, potential AEs such as acne, adults (1,2). Despite the established role of glucocorticoids weight gain, and sleep/mood disturbance may be mild. For in controlling short-term inflammation, and despite others, perceived AEs such as cataracts or glucocorticoid-induced osteoporosis with subsequent fracture may be Supported in part by the Agency for Healthcare Research
severe and result in substantial morbidity and mortality and Quality grant HS10389, the National Institute for Ar-
(7–15). Estimates regarding the prevalence and severity of thritis and Musculoskeletal and Skin Diseases grant P60-
many glucocorticoid-associated AEs are largely unknown, AR-48095, and NIH grant T32-AR-47512-03.
1Jeffrey R. Curtis, MD, MPH, Andrew O. Westfall, MS,
especially those less likely to require hospitalization or Jeroan Allison, MD, MS, Varghese George, PhD, Kenneth G.
specific medications for prevention or treatment. Of note, Saag, MD, MSc: University of Alabama at Birmingham; 2Jo-
the relationship between these common AEs and various hannes W. Bijlsma, MD: University Medical Center, Utrecht,
The Netherlands; 3Allison Freeman, MS, Claire M. Spettell,
glucocorticoid doses is also unclear, and the safety of PhD: Aetna Integrated Informatics, Aetna, Inc., Blue Bell,
long-term, low-dose prednisone (e.g., daily dosage Յ7.5 Pennsylvania; 4Stacey H. Kovac, PhD: Durham VA Medical
Center, and Duke University, Durham, North Carolina.
We hypothesized that the prevalence of AEs associated Address correspondence to Kenneth G. Saag, MD, MSc,
University of Alabama at Birmingham, Division of Clinical
with even low-dose glucocorticoid use would be high and Immunology and Rheumatology, FOT 820, 510 20th Street
would be dose and duration dependent. We therefore con- South, Birmingham, AL 35294. E-mail: [email protected].
ducted a population-based survey of glucocorticoid users Submitted for publication May 16, 2005; accepted in re-
to obtain prevalence estimates of glucocorticoid-associ- vised form August 18, 2005.
ated AEs. We specifically focused on these endpoints Glucocorticoids and Adverse Events among individuals who had prolonged use of an average Glucocorticoid survey administration. A 4-page, 69-
daily dose of Յ7.5 mg of prednisone.
question survey was mailed to 6,517 members who met theinclusion criteria for long-term glucocorticoid use. A fol-lowup postcard and a second survey were mailed to non- SUBJECTS AND METHODS
responders within 60 days of the first mailing. In thesurvey, we asked individuals to confirm that they had Data sources. Long-term glucocorticoid users were
actually taken oral glucocorticoids and whether they had identified within a national managed care organization experienced Ն1 of 8 specific AEs commonly attributed to (MCO) population covering 3 million lives in 36 states.
glucocorticoid use (sleep disturbance, acne, skin bruising/ These users and some of their demographic and disease thinning, weight gain, mood problems, diabetes or high characteristics were identified from the linked claims and blood sugar, cataracts, and fractures) while taking oral pharmacy databases of the MCO’s health maintenance or- glucocorticoids. Respondents were asked to classify symp- ganization, point of service, and preferred provider orga- toms as not, a little, or very bothersome. Individuals were nization health plans. Diseases, medications, and proce- not requested to speculate on a causal relationship be- dures were identified using International Classification of tween glucocorticoid use and the symptoms listed, but Diseases, Ninth Revision codes; National Drug Codes; and were only asked, “Have you ever had . . . while taking Common Procedural Terminology codes, respectively. All steroids?” Subjects were also asked whether their physi- individuals in the cohort had pharmacy benefits with vari- cian had discussed possible AEs of glucocorticoid use with them prior to use. Questions left blank or with mul-tiple responses were treated as missing data and were Cohort inclusion criteria. Individuals were character-
ized as long-term glucocorticoid users if they were Ն18years of age and had filled Ն60 days of outpatient oral Statistical analyses. Descriptive statistics characterized
glucocorticoid prescriptions from July 1, 2001 to Decem- basic demographics of survey responders and nonre- ber 31, 2002. Each member was enrolled in the health plan sponders. Categorical variables were compared using the at least 6 months prior to and following the first qualifying chi-square test of independence, and continuous variables glucocorticoid prescription (the index date). A total of were compared using the 2-tailed t-test. Despite MCO 6,517 health plan members met the inclusion criteria for pharmacy data documenting filled prescriptions for at long-term glucocorticoid use and were selected for further least 60 days of glucocorticoids for each subject, only those who confirmed that they had actually taken the glucocor- Because of potential differences in the prevalence of ticoids were analyzed. To facilitate comparison of AEs both short- and long-term AEs associated with incident related to glucocorticoid dose, cumulative dose (over 24 and prevalent glucocorticoid use, individuals were de- months) was categorized into quartiles (Ͻ1.7 gm, 1.7–2.8 scribed as new or prevalent users. New users were defined gm, 2.8 – 4.7 gm, and Ͼ4.7 gm of prednisone) and average as adults who had not received any glucocorticoid pre- daily dose was categorized into groups of clinical rele- scription in the 90 days prior to their index date; all others vance (0 –7.5 mg, 7.6 –20 mg, and Ͼ20 mg of prednisone were considered prevalent users. Individuals were ex- cluded from the study if they had diseases that might Multivariable stepwise logistic regression was per- impact outcomes of interest. These diagnoses included formed to identify significant associations between cumu- human immunodeficiency virus infection, alcoholism, de- lative glucocorticoid dose and AEs. Cumulative glucocor- mentia, lymphoma, leukemia, history of organ transplan- ticoid dose was included in all models, and age and sex tation, Paget’s disease of bone, and local or metastatic solid were also included in the fracture models. To test hypoth- eses related to the safety of increasing duration of useamong recipients of Յ7.5 mg/day of prednisone, total du- Glucocorticoid
medications. Qualifying
ration of use was forced into each model. For all other coids identified using pharmacy data included pred- variables, a univariate P value Ͻ 0.25 was required for nisone, prednisolone, dexamethasone, and other oral sys- entry into the model and a P value Ͻ 0.05 was required to temic glucocorticoids commonly used in practice in the remain in the model. Model building was conducted ac- United States. Injectable, topical, and intraocular glu- cording to Hosmer and Lemeshow (16), and model dis- cocorticoids were not included. Each prescription was crimination was assessed using the c statistic (17). To converted to a daily measure of a prednisone-equivalent assess trends across glucocorticoid dose groups, we mod- dose (in milligrams). The number of days the individual eled increasing dose category as a continuous variable and received oral glucocorticoids within the 18-month medi- used a Wald statistic to assess significance. All analyses cation eligibility window was examined to calculate the were performed using SAS software (SAS Institute, Cary, average and cumulative glucocorticoid exposure. Prescrip- tions for implausible daily doses (i.e., Ͼ100 mg/day oforal prednisone) were recoded using the patient’s averagedaily dose computed without these outlier prescriptions.
Of the 60,465 total glucocorticoid prescriptions, only 213(0.35%) prescriptions among 134 subjects needed to be A total of 2,446 (38%) managed care enrollees that met the definition for long-term glucocorticoid use returned the ticoids, 7% said that they had never taken them, and 2% Table 1. Characteristics of long-term glucocorticoid
did not answer (data not shown). Sixty-five percent of the (GC) users responding to survey (n ؍ 2,167)*
ever users reported current use at the time they completed Demographics
the survey. Among those who reported ever using glu-cocorticoids, a majority (68%) reported discussing poten- tial glucocorticoid-related AEs with their doctor prior to Compared with survey responders, the 4,071 survey nonresponders were younger (mean age 48 versus 53 years) and were more likely to be male (36% versus 30%).
Nonresponders’ duration of enrollment in the health plan, mean prednisone dose, and average number of comorbid conditions were similar to survey responders (data not Relationship between glucocorticoid dose and AEs.
The prevalence and self-reported severity of AEs experi- enced, stratified by quartile of cumulative glucocorticoid use, are shown in Figure 1. The AE with the greatest self-reported prevalence was weight gain, which was ex- perienced by almost 80% of subjects in the highest quartile of glucocorticoid use. Skin bruising/thinning and sleep disturbance were the next most commonly reported AEs.
Cataracts (15% overall) and fractures (12% overall) were reported less frequently. Only 10% of subjects reported that they had not experienced any of the AEs concurrent with glucocorticoid use. We examined claims data for fractures separately during the 30-month observation pe- riod. Ten percent of the cohort had Ն1 medical service claims for a fracture during the 2.5-year period of obser- The adjusted relative risks of AEs associated with cu- mulative glucocorticoid dose are described in Table 2. A strong dose-response relationship was observed between increasing quartiles of glucocorticoid use and all of the AEs examined. To assess the validity of our self-reported * N ϭ 2,167 subjects after excluding individuals who did not con-firm that they had actually taken glucocorticoids. Values are thenumber (percentage) unless otherwise indicated.
† Total Ͻ2,167 because of missing survey data.
‡ During study period of 2 years.
survey. Their characteristics are described in Table 1. In-dividuals were middle aged, predominantly women, andreceived prescriptions for an average of 16 mg/day of pred-nisone. A majority were new users rather than prevalentglucocorticoid users. Persons with rheumatoid arthritiswere prescribed a mean Ϯ SD dosage of 12 Ϯ 8 mg/day ofprednisone, in contrast to persons with inflammatory Figure 1. Prevalence of adverse events associated with long-term
bowel disease who received a mean Ϯ SD prednisone glucocorticoid use stratified by cumulative glucocorticoid dosage(n ϭ 2,167 subjects after excluding individuals who did not con- dosage of 31 Ϯ 16 mg/day (data not shown). The mean firm that they had actually taken glucocorticoids). Quartiles of daily prednisone doses for the other diseases, including cumulative prednisone-equivalent glucocorticoid dosage: Q1 ϭ systemic lupus erythematosus, chronic obstructive pulmo- Ͻ1.7 gm (e.g., 10 mg/day of prednisone for 6 months); Q2 ϭ nary disease, and asthma, fell between these 2 extremes.
1.7–2.8 gm (e.g., 10 mg/day of prednisone for 9 months);Q3 ϭ 2.9 – 4.7 gm (e.g., 10 mg/day of prednisone for 12 months); Hypertension and rheumatoid arthritis were among the Q4 ϭ Ͼ4.7 gm (e.g., 10 mg/day of prednisone for 18 months).
most common comorbid diagnoses. More than 91% of * P Ͻ 0.0001 (Cochran-Armitage trend test). † P Ͻ 0.01 (Cochran- respondents indicated that they had ever taken glucocor- Armitage trend test). ‡ P Ͻ 0.001 (Cochran-Armitage trend test).
Glucocorticoids and Adverse Events fracture models, we separately modeled fractures using Table 2. Relationship between cumulative prednisone-
medical claims data. Even after multivariable adjustment, equivalent dose and self-reported adverse events*
a positive and significant association with increasing cu- Adjusted
mulative glucocorticoid dose was observed (data not Characteristics/quartile†
Among users of Ͼ7.5 mg/day of prednisone, glucocorti- coid dose and duration of use were strongly associated with all AEs (data not shown). The adjusted odds ratios of increasing duration of glucocorticoid use among individ- uals prescribed Յ7.5 mg/day of prednisone are presented in Table 3. Acne, skin bruising, weight gain, and cataracts were significantly associated with longer durations of low- dose glucocorticoid use. In contrast, increasing daily dose (within the 0 –7.5 mg/day range) was more strongly asso- ciated with sleep disturbance and fractures than was in- creased duration of use. The Spearman correlation coeffi- cient between cumulative glucocorticoid dose and average DISCUSSION
We found a high prevalence of self-reported AEs associ- ated with glucocorticoid use among individuals treated long-term for a variety of conditions. Almost all respon- dents reported at least 1 AE that was temporally associated with concurrent glucocorticoid use, and more than half reported that at least 1 AE was very bothersome. Serious AEs potentially attributable to glucocorticoid exposure were also common, with 15% of the cohort self reporting cataracts and 12% self reporting Ն1 fractures. Increasing cumulative glucocorticoid dose had a strong positive as- sociation with serious AEs typically attributed to gluco- corticoid use, even after multivariable adjustment for demographic and disease comorbidities that were also as- Long-term observational studies of patients receiving commonly prescribed doses of glucocorticoids have re- ported high rates of glucocorticoid-associated AEs. A study of 120 patients with giant cell arteritis (mean age at diagnosis 75 years) followed for a median of 10 years found that 86% of patients experienced Ն1 AEs (18); 58% of the patients experienced Ն2 serious AEs, the most com- mon of which were cataracts (41%) and fractures (38%). Ina similarly designed study of 232 patients with polymyal- * Total sample size for each model shown in parentheses less thanentire cohort (n ϭ 2,167) due to missing survey responses. OR ϭ gia rheumatica (19) (mean age at diagnosis 73 years) fol- odds ratio; 95% CI ϭ 95% confidence interval.
lowed for an average of 8 years, 65% of individuals treated † Quartiles of cumulative glucocorticoid dosage (prednisone equiv- with glucocorticoids experienced at least 1 AE, the most alents): Q1 ϭ Ͻ1.7 gm (e.g., 10 mg/day of prednisone for 6 months);Q2 ϭ 1.7–2.8 gm (e.g., 10 mg/day of prednisone for 9 months); common of which were vertebral fracture (18%) and cat- Q3 ϭ 2.9 – 4.7 gm (e.g., 10 mg/day of prednisone for 12 months); aracts (36%). Cumulative glucocorticoid dose had a strong Q4 ϭ Ͼ4.7 gm (e.g., 10 mg/day of prednisone for 18 months).
association with AEs, particularly after the dosage reached ‡ Adjusted for age, sex, and number of comorbid diseases; c statis-tic ϭ 0.66.
1.8 gm, which was the approximate cutpoint between the § Adjusted for age, sex, and ethnicity; c statistic ϭ 0.76.
first and second quartiles in our analyses. Our results are ¶ Adjusted for age, sex, new user, and number of comorbid dis-eases; c statistic ϭ 0.67.
similar to these studies because 90% of our subjects re- # Adjusted for age, sex, new user, ethnicity, and number of comor- ported at least 1 AE while receiving glucocorticoids. Rates of self-reported fracture (12%) and cataract (15%) in our ** Adjusted for age, sex, ethnicity, and number of comorbid dis-eases; c statistic ϭ 0.69.
study were high, although lower than these 2 reports, †† Adjusted for ethnicity and number of comorbid diseases; c sta- likely reflecting a younger population, diverse indications for glucocorticoid treatment, and lower cumulative glu- ‡‡ Adjusted for age, ethnicity, number of comorbid diseases, andincome; c statistic ϭ 0.72.
§§ Adjusted for age, sex, income, and number of comorbid diseases; More recent interest has focused on the clinical efficacy and potential disease-modifying properties of long-term Table 3. Relationship between glucocorticoid dose and duration and adverse events among low-dose (prednisone
<7.5 mg/day) glucocorticoid users (n ؍ 670)*
Average daily dose
90-day increase in
(per 1-mg/day increase
Characteristics
duration of use
within 0–7.5 mg/day range)
* 95% CI ϭ 95% confidence interval; NS ϭ not significant.
† Adjusted for age, sex, and number of comorbid diseases; c statistic ϭ 0.63.
‡ Adjusted for age and sex; c statistic ϭ 0.77.
§ Adjusted for age and sex; c statistic ϭ 0.65.
¶ Adjusted for age and sex; c statistic ϭ 0.68.
# Adjusted for age, ethnicity, and heart failure; c statistic ϭ 0.69.
** Adjusted for ethnicity; c statistic ϭ 0.64.
†† Adjusted for age and number of comorbid diseases; c statistic ϭ 0.71.
‡‡ Adjusted for age, sex, income, and number of comorbid diseases; c statistic ϭ 0.65.
low-dose prednisone (Յ7.5 mg/day), particularly for pa- mg/day over 24 months. These data suggest that long-term tients with arthritis. In small clinical trials of patients with use of low-dose prednisone in a large, diverse managed rheumatoid arthritis randomized to low glucocorticoid care population is typically administered only to a minor- doses (prednisolone 7.5–10 mg daily), vertebral fractures, ity of individuals with rheumatoid arthritis.
weight gain, and hyperglycemia were observed among the Our study has several strengths. The large, population- glucocorticoid-treated patients, although incidence rates based design coupled with the claims data and pharmacy of these AEs were low (3,4). These and other trials with linkage permitted us to analytically account for a variety of aggressive glucocorticoid tapering regimens (5) have dem- comorbid illnesses that were ascertained using both self- onstrated that low-dose or rapidly tapered glucocorticoids reported and administrative sources. Pharmacy data, may result in minimal short-term glucocorticoid-associ- rather than self report, were used to determine glucocorti- ated toxicity. The strengths of these randomized, prospec- coid use patterns and thus reduce potential recall bias. We tive, placebo-controlled trials are their blinded outcomes were able to capture self-reported AEs that are often not assessment; lack of confounding by disease severity; and assessed in observational studies that utilize only admin- prohibition of even short-term, high-dose glucocorticoid istrative data or even medical chart review. Our study exposure. Factors that may limit their generalizability in- captured the diverse diseases treated with glucocorticoids clude selection of individuals with few comorbid illnesses and included use of intermittent glucocorticoid prescrip- and insufficient duration of followup to detect many of the tions, allowing better generalizability than other studies AEs that develop and/or progress over time.
that have focused on populations treated with glucocorti- In contrast to these clinical trial findings, an observa- coids for only 1 condition or with only continuous ther- tional study with a longer followup period compared pa- apy. Finally, we were able to examine demographic covari- tients with rheumatoid arthritis treated with a mean of 7 ates such as ethnicity and education that are rarely mg/day of near-continuous prednisone therapy for 4.9 available in medical record or claims-based data sources.
years with similar patients with rheumatoid arthritis not One limitation of our work is that AEs were self re- treated with glucocorticoids (12). In the glucocorticoid- ported, and confirmation by a physician was not available.
treated group, 82% experienced at least 1 AE compared Although self report may be the only reasonable source of with 24% in the control group. Fractures (19%) and cata- information for certain subjective AEs such as mood and racts (15%) were the most common serious AEs. We also sleep disturbance, acne, and skin changes associated with found several AEs that were significantly associated with glucocorticoid use, it may not be optimal for other out- longer durations and small increments in glucocorticoid comes such as fracture. However, the positive predictive dose, even among Յ7.5 mg/day prednisone users (Table 3), value of self-reported fracture has been reported to be and we demonstrated a dose-dependent increase in frac- Ͼ80% at any fracture site and often in excess of 95% for ture risk in this group that is concordant with other studies osteoporotic fracture sites (21–24). Because of uncertainty of low-dose glucocorticoid users (10,20). Moreover, al- as to whether a medical claim for fracture represented a though the 892 individuals with rheumatoid arthritis in new (incident) fracture or followup for a prevalent frac- our study received the lowest mean daily dosage of pred- ture, we censored fracture data after a single event and nisone (12 mg/day) compared with those with other dis- therefore likely underestimated fracture incidence among eases that we examined, only 40% received a mean of Յ7.5 individuals who experienced Ͼ1 new fracture. Vertebral Glucocorticoids and Adverse Events fractures may be asymptomatic and may not be identified uals the expected benefits will outweigh the possible risks.
in as many as two-thirds of patients (25,26), which pro- For example, among patients with early rheumatoid arthri- vides an additional reason for our fracture estimates to be tis, combination therapy that includes glucocorticoids conservative. We were able to compare our results based may have a sustained, disease-modifying effect (30). How- on self-reported outcomes with fracture outcomes from ever, minimizing potentially avoidable AEs such as glu- administrative data and achieved similar results. Com- cocorticoid-induced osteoporosis, for which there are effi- pared with fractures, less research has been conducted to cacious but underutilized therapies (31–34), must be a validate self report of cataracts, but the positive predictive shared responsibility among physicians and patients. De- value of self-reported cataracts has been estimated to be spite advances in chronic disease management, it is un- ϳ83% (95% confidence interval 79–88%) (27).
likely that the long-term use of glucocorticoids will disap- Another limitation of our work is that the self-reported pear any time soon. As such, heightened attention to outcome data are cross-sectional, and determining the common glucocorticoid-associated AEs that bother pa- temporal sequence of AEs to implicate a causal relation- tients, even among users of lower doses, is warranted.
ship with glucocorticoid exposure is problematic. For thisreason, we intentionally excluded persons with medically REFERENCES
recognized diabetes from our high blood sugar analysis(16% of the cohort). This conservative strategy underesti- 1. Van Staa TP, Leufkens HG, Abenhaim L, Begaud B, Zhang B, mates the proportion of those who developed incident Cooper C. Use of oral corticosteroids in the United Kingdom.
QJM 2000;93:105–11.
diabetes as a consequence of glucocorticoid therapy. De- 2. Walsh LJ, Wong CA, Pringle M, Tattersfield AE. Use of oral spite the similarity of survey respondents and nonrespon- corticosteroids in the community and the prevention of sec- dents in the distribution of comorbid conditions and mean ondary osteoporosis: a cross sectional study. BMJ 1996;313: daily prednisone dose (ascertained using administrative 3. Kirwan JR, and the Arthritis and Rheumatism Council Low- data), possible response bias may have influenced our Dose Glucocorticoid Study Group. The effect of glucocorti- findings if individuals that experienced glucocorticoid- coids on joint destruction in rheumatoid arthritis. N Engl associated AEs were more likely to respond to the survey.
Additionally, glucocorticoids have been associated with a 4. Van Everdingen AA, Jacobs JW, Siewertsz van Reesema DR, variety of other AEs (e.g., infection, hypertension) that we Bijlsma JW. Low-dose prednisone therapy for patients withearly active rheumatoid arthritis: clinical efficacy, disease- modifying properties, and side effects: a randomized, double- Finally, we were not able to compare our prevalence of blind, placebo-controlled clinical trial. Ann Intern Med 2002; self-reported AEs with that of a control group with similar diseases of equal severity that did not receive glucocorti- 5. Boers M, Verhoeven AC, Markusse HM, van de Laar MA, Westhovens R, van Denderen JC, et al. Randomised compar- coids. Confounding by disease severity is a concern in ison of combined step-down prednisolone, methotrexate and observational studies where “sicker” patients have in- sulphasalazine with sulphasalazine alone in early rheumatoid creased exposure to a drug of interest (28). We adjusted for arthritis [published erratum appears in Lancet 1998;351:220].
general comorbidity using the number of unique diseases for which the individual received medical care in the prior 6. Hansen M, Podenphant J, Florescu A, Stoltenberg M, Borch A, Kluger E, et al. A randomised trial of differentiated pred- 30 months. However, because of the diverse nature of the nisolone treatment in active rheumatoid arthritis: clinical diseases for which individuals were prescribed glucocor- benefits and skeletal side effects. Ann Rheum Dis 1999;58: ticoids, we were not able to adjust for all factors indicative of more severe disease (e.g., rheumatoid nodules and ero- 7. National Osteoporosis Foundation. Physician’s guide to pre- vention and treatment of osteoporosis. Belle Mead (NJ): Ex- sions for patients with rheumatoid arthritis); therefore, residual confounding by disease indication or severity is 8. Skalka HW, Prchal JT. Effect of corticosteroids on cataract possible. However, we demonstrated a strong dose-re- formation. Arch Ophthalmol 1980;98:1773–7.
sponse relationship between rates and severity of AEs and 9. Van Staa TP, Abenhaim L, Cooper C, Zhang B, Leufkens HG.
increasing doses of cumulative and daily oral glucocorti- Public health impact of adverse bone effects of oral cortico-steroids. Br J Clin Pharmacol 2001;51:601–7.
coids, even after adjusting for a variety of demographic 10. Van Staa TP, Leufkens HG, Abenhaim L, Zhang B, Cooper C.
Use of oral corticosteroids and risk of fractures. J Bone Miner The prevalence of overall and very bothersome AEs associated with long-term glucocorticoid use was high, 11. Van Staa TP, Leufkens HG, Cooper C. The epidemiology of corticosteroid-induced osteoporosis: a meta-analysis. Osteo- even among users of lower doses, and this observation may be underappreciated by clinicians. Because patients may 12. Saag KG, Koehnke R, Caldwell JR, Brasington R, Burmeister be concerned about the potential AEs of glucocorticoids LF, Zimmerman B, et al. Low dose long-term corticosteroid prior to use (29), physicians need to be well apprised of therapy in rheumatoid arthritis: an analysis of serious adverse their risks. A majority of individuals in our cohort re- 13. Michel BA, Bloch DA, Fries JF. Predictors of fractures in early ported discussing potential glucocorticoid-associated AEs rheumatoid arthritis. J Rheumatol 1991;18:804 – 8.
with their physician prior to initiating therapy. This ob- 14. Merlino LA, Bagchi I, Taylor TN, Utrie P, Chrischilles E, servation highlights the opportunity for physicians to dis- Sumner W 2nd, et al. Preference for fractures and other glu- cuss with patients the possible AEs of glucocorticoids in cocorticoid-associated adverse effects among rheumatoid ar-thritis patients. Med Decis Making 2001;21:122–32.
the context of their expected benefits.
15. Klein BE, Klein R, Lee KE, Danforth LG. Drug use and five- Although the best way to reduce glucocorticoid-associ- year incidence of age-related cataracts: the Beaver Dam Eye ated AEs is to limit exposure and dose, for some individ- Study. Ophthalmology 2001;108:1670 – 4.
16. Hosmer JD, Lemeshow S. Applied logistic regression. 2nd ed.
26. Kado DM, Browner WS, Palermo L, Nevitt MC, Genant HK, Hoboken (NJ): John Wiley and Sons; 2000.
Cummings SR, and the of Osteoporotic Fractures Research 17. Lemeshow S, Hosmer DW Jr. A review of goodness of fit Group. Vertebral fractures and mortality in older women: a statistics for use in the development of logistic regression prospective study. Arch Int Med 1999;159:1215–20.
models. Am J Epidemiol 1982;115:92–106.
27. Bergmann MM, Byers T, Freedman DS, Mokdad A. Validity of 18. Proven A, Gabriel SE, Orces C, O’Fallon W, Hunder G. Glu- self-reported diagnoses leading to hospitalization: a compar- cocorticoid therapy in giant cell arteritis: duration and ad- ison of self-reports with hospital records in a prospective verse outcomes. Arthritis Rheum 2003;49:703– 8.
study of American adults. Am J Epidemiol 1998;147:969 –77.
19. Gabriel SE, Sunku J, Salvarani C, O’Fallon WM, Hunder GG.
28. Salas M, Hofman A, Stricker BH. Confounding by indication: Adverse outcomes of antiinflammatory therapy among pa- an example of variation in the use of epidemiologic terminol- tients with polymyalgia rheumatica. Arthritis Rheum 1997; ogy. Am J Epidemiol 1999;149:981–3.
29. Morrison E, Crosbie D, Capell HA. Attitude of rheumatoid 20. Ton FN, Gunawardene SC, Lee H, Neer RM. Effects of low- arthritis patients to treatment with oral corticosteroids. Rheu- dose prednisone on bone metabolism. J Bone Miner Res 2005; matology (Oxford) 2003;42:1247–50.
30. Landewe RB, Boers M, Verhoeven AC, Westhovens R, van de 21. Honkanen K, Honkanen R, Heikkinen L, Kroger H, Saarikoski Laar MA, Markusse HM, et al. COBRA combination therapy in S. Validity of self-reports of fractures in perimenopausal patients with early rheumatoid arthritis: long-term structural women. Am J Epidemiol 1999;150:511– 6.
benefits of a brief intervention. Arthritis Rheum 2002;46:347– 22. Nevitt M, Cummings SR, Browner WS, Seeley DG, Cauley JA, Vogt TM, et al. The accuracy of self-report of fractures in 31. Curtis JR, Westfall AO, Allison JJ, Becker A, Casebeer L, elderly women: evidence from a prospective study. Am J Freeman A, et al. Longitudinal patterns in the prevention of osteoporosis in glucocorticoid-treated patients. Arthritis 23. Ivers RQ, Cumming RG, Mitchell P, Peduto AJ. The accuracy of self-reported fractures in older people. J Clin Epidemiol 32. Yood RA, Harrold LR, Fish L, Cernieux J, Emani S, Conboy E, et al. Prevention of glucocorticoid-induced osteoporosis: ex- 24. Ismail AA, O’Neill TW, Cockerill W, Finn JD, Cannata JB, perience in a managed care setting. Arch Intern Med 2001; Hoszowski K, et al, and the European Prospective Osteoporo- sis Study Group. Validity of self-report of fractures: results 33. Solomon DH, Katz JN, Jacobs JP, la Tourette AM, Coblyn J.
from a prospective study in men and women across Europe.
Management of glucocorticoid-induced osteoporosis in pa- tients with rheumatoid arthritis: rates and predictors of care 25. Vogt TM, Ross PD, Palermo L, Musliner T, Genant HK, Black in an academic rheumatology practice. Arthritis Rheum 2002; D, et al, and the Fracture Intervention Trial Research Group.
Vertebral fracture prevalence among women screened for the 34. Chantler IW, Davie MW, Evans SF, Rees JS. Oral corticoste- Fracture Intervention Trial and a simple clinical tool to screen roid prescribing in women over 50, use of fracture prevention for undiagnosed vertebral fractures. Mayo Clin Proc 2000;75: therapy, and bone densitometry service. Ann Rheum Dis

Source: http://andy.x10.bz/Pubs/Pub023.pdf