Key Points
Question
Has the frequency of potassium level monitoring among healthy young women treated for acne with spironolactone decreased following recent study findings suggesting its low usefulness and updated guideline recommendations for this population?
Findings
In this cohort study, the overall proportion of 108 547 women with potassium level monitoring decreased from 41% in 2008 to 38% in 2018, with a greater decrease observed among women treated by dermatologists (49% to 41%) than by internists (40% to 38%) but an increase (71% to 75%) among women treated by advanced practice clinicians (ie, nurse practitioners and physician assistants).
Meaning
This study finds that potassium level monitoring remains common among young women treated for acne with spironolactone, suggesting a need for future efforts to address this practice gap.
Abstract
Importance
Although recent studies and guideline recommendations indicate that potassium level monitoring is of low usefulness for healthy young women being treated for acne with spironolactone, little is known about whether these recommendations have been implemented in clinical practice.
Objective
To evaluate trends in rates of potassium level monitoring among young women treated for acne with spironolactone and clinician variability in monitoring practices.
Design, Setting, and Participants
This retrospective cohort study was conducted between January 1, 2008, and June 30, 2019, using data from the Optum deidentified Clinformatics Data Mart database. Participants comprised 108 547 healthy female patients aged 12 to 45 years who were treated for acne with spironolactone.
Main Outcome and Measures
The primary outcome was the proportion of women who received a test for baseline potassium level and the proportion of women whose potassium levels were monitored in the first 180 days of being treated for acne with spironolactone. For clinicians who had prescribed at least 5 courses of spironolactone, the percentage of clinicians who ordered baseline potassium testing or monitoring and the percentage of clinicians who always monitored potassium were compared between 2008 and 2015 vs between 2016 and 2018.
Results
For 108 547 women included in this cohort study, the mean (SD) age at the start of treatment was 30.7 (8.6) years, and the mean (SD) course duration was 159 (218) days. Between 2008 and 2018, the percentage of women whose potassium levels were monitored within 180 days of starting spironolactone by all clinicians decreased from 41.4% to 38.5%, with a decrease from 48.9% to 41.0% among dermatologists and from 39.7% to 37.7% among internists but with an increase from 71.4% to 75.4% among advanced practice clinicians (ie, nurse practitioners and physician assistants). The proportion of dermatologists who always monitored potassium decreased from 10.6% between 2008 and 2015 to 4.2% between 2016 and 2018. There was no significant difference in the proportion of internists who always monitored potassium from 2008 to 2015 (15.8%) vs from 2016 to 2018 (17.7%).
Conclusions and Relevance
Despite increasing evidence and guidelines supporting the elimination of potassium monitoring among healthy young women treated for acne with spironolactone, the present study findings suggest that potassium monitoring remains common, with substantial variability in clinician practices. There is a need for future implementation and dissemination research to understand underlying factors for this variation and to develop strategies to address this practice gap.
This cohort study assesses the change in the rates of potassium level monitoring from 2008 to 2018 by clinicians overall and stratified by dermatologists, internists, nurse practitioners, and physician assistants in healthy young women treated for acne with spironolactone.
Introduction
Spironolactone is an effective treatment of acne in women, and its use is growing over time.1,2 Although concerns around risk of hyperkalemia have been raised, 2 large observational studies have mitigated this concern and have suggested low value in potassium monitoring of healthy young women being treated for acne with spironolactone.3,4 In addition, no increased risk among patients treated with spironolactone and drospirenone-containing combined oral contraceptives has been identified.5,6 These studies have led to a change in guidelines and literature redefining standard practice for laboratory monitoring. Updated guidelines on spironolactone that support eliminating potassium monitoring for healthy young women being treated for acne were adopted by popular clinical references, such as UpToDate, and by the American Academy of Dermatology in 2016.7,8,9
Despite this evolving landscape, little is known about how changes in understanding of the value of potassium monitoring in individuals treated with spironolactone have impacted clinical practice. The purpose of the present study was to identify temporal trends in potassium monitoring for women treated with spironolactone and to evaluate clinician variability in potassium-monitoring practices.
Methods
This study was a retrospective cohort study of patients in the Optum deidentified Clinformatics Data Mart database from January 1, 2008, through June 30, 2019. The Optum Clinformatics Data Mart includes deidentified commercial claims data for approximately 12 to 14 million individuals annually in the United States. These data include both medical and pharmacy claims as well as patient demographic information, such as age and sex. The patient population available in the Optum Clinformatics Data Mart is similar to the demographic characteristics of the United States population with respect to sex, age, and geographic distribution.10 This study was conducted in adherence with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.11 The University of Pennsylvania deemed this study to be exempt from review by the institutional review board and waived the requirement for obtaining informed consent because the study involved deidentified data.
Study Population
Inclusion criteria were (1) female sex; (2) aged 12 to 45 years; (3) at least 1 encounter with an International Classification of Diseases, Ninth Revision (ICD-9) or International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10) codes for acne; (4) prescription for a course of spironolactone of at least 30 days in duration that was started within 30 days of this encounter for acne, with the start date of the course being defined as the index date; and (5) at least 180 days of continuous enrollment in Optum Clinformatics Data Mart prior to and after the index date. Prior studies have validated the use of ICD-9 and ICD-10 codes to identify patients with acne.12,13 To account for logistical considerations and nonadherence that could result in delays between prescriptions, prescriptions separated by fewer than 30 days from the date of the prior prescription plus the number of days of medication supplied were considered to be part of the same course of therapy.14,15,16,17,18,19,20
The primary outcome was the frequency of testing baseline potassium levels (defined as potassium testing between the encounter for acne and the index date) and the frequency of monitoring potassium levels while being treated (defined as testing ordered within 180 days of index date). Claims for potassium testing were identified by Logical Observation Identifiers Names and Codes. In addition, for clinicians who had prescribed at least 5 courses of spironolactone, we evaluated the frequency of baseline potassium testing and potassium monitoring within 180 days of the index date. To assess whether clinical evidence was associated with clinician variation in potassium-monitoring practices, we compared the frequency of clinicians who always tested baseline potassium levels and monitored potassium levels during treatment for their patients between the period before the Plovanich et al3 study and the updated American Academy of Dermatology guidelines (2008-2015) with the period after these recommendations were published (2016-2018).8
In addition, to examine whether potassium monitoring was associated with treatment management, we evaluated the frequency with which patients had a clinical encounter with an ICD-9 or ICD-10 code for hyperkalemia (276.7, E875) and how often spironolactone was discontinued within 30 days of this encounter.
Statistical Analysis
To evaluate for changes in potassium monitoring over time, an interrupted time series analysis was conducted using a linear regression model with an inflection point separating the periods from 2008 to 2015 vs from 2016 to 2018. Clinician variability between the periods from 2008 to 2015 vs from 2016 to 2018 were compared using χ2 tests. Subgroup analyses were performed among dermatologists, internists, and advanced practice clinicians (ie, nurse practitioners and physician assistants). Clinician specialty was classified using taxonomy codes associated with the clinical encounter claims. Because clinician specialty classification data were incomplete, we conducted a sensitivity analysis in which we excluded courses prescribed by clinicians with missing specialty classification data. Multivariable logistic regression was used to evaluate for a potential association between volume of spironolactone use and the frequency of clinicians who always tested baseline potassium levels and monitored potassium levels during treatment, controlling for patient age and stratified by clinician specialty. Statistical analyses were performed in Stata, version 15 (StataCorp LLC). A 2-sided value of P < .05 was considered statistically significant.
Results
Among 108 547 patients meeting the inclusion criteria, the mean (SD) age at the start of treatment was 30.7 (8.6) years (Table). The mean (SD) course duration was 159 (218) days. Of the spironolactone courses, 50.3% were prescribed by dermatologists, 23.9% by internists, and 0.6% by advanced practice clinicians (specialty data were unavailable for 25.3% of courses). Most advanced practice clinicians (79.7%) were associated with a dermatology practice classification. The use of spironolactone increased over time, with 6639 courses started in 2008 (3.0 courses per 100 women with encounters for acne) compared with 16 895 in 2018 (7.2 courses per 100 women with encounters for acne). In our cohort, 0.22% of patients had a clinical encounter with a diagnosis code for hyperkalemia during the course of spironolactone treatment, and 0.03% discontinued spironolactone within 30 days of a clinical encounter with a diagnosis code for hyperkalemia.
Table. Demographic Characteristics of Patients.
| Characteristic | Spironolactone prescriber specialty, % | |||
|---|---|---|---|---|
| Overall | Dermatologist | Internist | Advanced practice clinician | |
| Total No. | 108 547 | 54 583 | 25 935 | 610 |
| Age, mean (SD), y | 30.7 (8.6) | 29.4 (8.5) | 33.0 (8.2) | 29.9 (8.1) |
| Age group, y | ||||
| 12-19 | 11.9 | 14.6 | 6.4 | 9.5 |
| 20-30 | 36.8 | 40.2 | 30.9 | 44.2 |
| 31-40 | 35.2 | 32.6 | 39.9 | 33.3 |
| 41-45 | 16.1 | 12.6 | 22.8 | 13.0 |
| Course duration, mean (SD), d | 159 (218) | 158 (209) | 186 (264) | 283 (304) |
Between 2008 and 2018, testing for baseline potassium level decreased from 48.5% to 41.8% among all clinicians, with a decrease from 42.6% to 40.8% among dermatologists and from 50.2% to 42.7% among internists. By contrast, testing for baseline potassium level increased from 21.4% to 24.2% among advanced practice clinicians (Figure 1).
Figure 1. Temporal Trends in Frequency of Testing for Baseline Potassium Levels Among Young Women Treated for Acne With Spironolactone.
APCs indicates advanced practice clinicians.
Potassium level monitoring performed within 180 days of the first prescription fill for spironolactone decreased from 41.4% to 38.5% among all clinicians, with a decrease among dermatologists from 48.9% to 41.0% and a decrease among internists from 39.7% to 37.7%. By contrast, monitoring among advanced practice clinicians increased from 71.4% to 75.4% (Figure 2). In the interrupted time series analysis, potassium level monitoring within 180 days decreased among dermatologists by 0.3% per year (95% CI, 0.1%-0.6%), with no significant change to the baseline rate of monitoring or slope before and after 2016. No significant differences were noted among internists or advanced practice clinicians. Similar findings were observed in our sensitivity analysis excluding those with missing specialty classification data (eFigures 1 and 2 in the Supplement).
Figure 2. Temporal Trends in Frequency of Monitoring of Potassium Levels Among Young Women Treated for Acne With Spironolactone.
APCs indicates advanced practice clinicians.
From 2008 to 2015 among 3093 dermatologists with at least 5 prescriptions for spironolactone, 10.6% always monitored potassium levels within 180 days after starting spironolactone. This frequency decreased to 4.2% from 2016 to 2018 (eFigure 3 in the Supplement; P < .001). From 2008 to 2015 among 752 internists with at least 5 prescriptions for spironolactone, 15.8% always monitored potassium levels within 180 days after starting spironolactone. This frequency remained similar (17.7%) from 2016 to 2018 (eFigure 4 in the Supplement; P = .37). Between 2008 and 2015, 55.0% of advanced practice clinicians always monitored potassium levels compared with 42.9% from 2016 to 2018 (P = .49; n = 34). There was substantial heterogeneity in the frequency of potassium level monitoring among dermatologists and internists (eFigures 3 and 4 in the Supplement).
After controlling for patient age, each additional 10 total courses of spironolactone prescribed by a clinician was associated with significantly decreased odds of always monitoring potassium levels among all clinicians (odds ratio, 0.65; 95% CI, 0.53-0.79) and among dermatologists (odds ratio, 0.57; 95% CI, 0.46-0.71) but not among internists (odds ratio, 0.92; 95% CI, 0.76-1.13) or advanced practice clinicians (odds ratio, 1.08; 95% CI, 0.49-2.39).
Discussion
A growing body of evidence suggests low value for potassium level monitoring among healthy young women treated for acne with spironolactone, and this evidence has been incorporated into treatment guidelines.3,4,7,8 Nevertheless, the present study findings indicated that both baseline and routine monitoring of potassium levels remained common in 2018 among dermatologists and internists, with more than 40% of patients having their potassium level assessed at baseline and with more than 35% of patients having their potassium levels monitored by all clinicians during the first 6 months of treatment with spironolactone. These findings highlight the well-documented challenges of implementing new evidence into clinical practice and the need for future implementation and dissemination research to address this practice gap.21,22
Among dermatologists, there was a modest decrease in baseline and ongoing monitoring, both with respect to the overall frequency of monitoring and with respect to the number of clinicians who always monitored potassium. Unlike dermatologists, internists did not display a decrease in clinician variability, and many internists continued to assess potassium levels nearly universally for their patients being treated for acne with spironolactone. Although sample size was limited, advanced practice clinicians had high use of potassium level monitoring.
There are several potential explanations for these findings. First, it is possible that a portion of dermatologists incorporated emerging evidence into their practice.7,8,9 Because much of the emerging literature on this topic has been published in dermatology journals, the lack of change in practice patterns among internists and advanced practice clinicians could reflect decreased awareness of these new data and a need for targeted educational interventions. However, because there is not a clear inflection point in the frequency of testing around 2016, when the study by Plovanich et al3 and updated guidelines were released, it is also possible that these modest changes in practice occurred irrespective of the introduction of new clinical evidence and guidelines to the published literature.8
Another possibility is that increased clinical experience and comfort prescribing spironolactone has influenced the frequency of potassium level monitoring. As the use of spironolactone has grown during the last decade, clinicians may have become more comfortable with not assessing potassium levels when prescribing spironolactone. Furthermore, the lower proportion of dermatologists who always monitored laboratory values compared with internists and with advanced practice clinicians could reflect faster growth in the use of spironolactone among dermatologists compared with nondermatologists.1,23 Because few advanced practice clinicians in the present study prescribed at least 5 courses of spironolactone, their lower comfort and experience with spironolactone may explain their particularly high use of potassium level monitoring. Ultimately, understanding differences between those who have and those who have not changed their practice could provide valuable insights to guide future implementation efforts to address the remaining substantial variation in practice patterns.
We found that encounters associated with a diagnosis code for hyperkalemia were uncommon among patients being treated with spironolactone in this cohort. Encounters with a diagnosis code for hyperkalemia were noted for only 0.22% of treated patients and were associated with treatment discontinuation in only 3 per 10 000. These findings are consistent with a prior study that reported approximately 0.7% of potassium tests in healthy young women with acne identified hyperkalemia, whether or not the patient was treated with spironolactone.3 In addition, the extremely low rate of spironolactone discontinuation following an encounter with a diagnosis code for hyperkalemia observed in our study would support that these encounters are representative of false-positive testing rather than representative of a complication from spironolactone use.
Limitations
This study should be interpreted in the context of its design. It is possible that some tests for baseline potassium levels may have been ordered prior to the encounter for acne; thus, our definition of testing for baseline potassium level may underestimate the true incidence of baseline potassium testing. However, even with this conservative definition, baseline testing was common. Although it is possible that some of the tests of potassium levels ordered after starting spironolactone may have been for other indications, given the study population of young women being treated for acne, it is most likely that these tests were ordered in association with spironolactone use, particularly among patients being treated by dermatologists. Because claims data do not include laboratory values for patients whose potassium levels were determined, we cannot directly evaluate the potassium levels of patients treated with spironolactone.
In addition, the ideal percentage of patients who should undergo potassium level monitoring while being treated with spironolactone is unknown; nevertheless, given the low usefulness of testing in healthy young women being treated for acne, it is likely that the observed high frequency of testing is not optimal. Although our overall sample size was large and is a strength of the study, there were few advanced practice clinicians identified, which limits the conclusions that can be drawn from subgroup analyses in this population and may explain some of the variability in monitoring practices observed among advanced practice clinicians.
Conclusions
Despite growing evidence and guidelines supporting the elimination of potassium monitoring among healthy young women treated for acne with spironolactone, potassium monitoring remains common, with substantial variability in clinician practices. There is a need for future implementation and dissemination research to understand underlying factors for this variation and to develop strategies to address this practice gap.
eFigure 1. Temporal Trends in Frequency of Baseline Potassium Labs Among Women Treated with Spironolactone for Acne
eFigure 2. Temporal Trends in Frequency of Monitoring Potassium Labs Among Women Treated with Spironolactone for Acne
eFigure 3. Variation in Baseline and Routine Potassium Monitoring Among Dermatologists Prior to and After 2016
eFigure 4. Variation in Baseline and Routine Potassium Monitoring Among Internists Prior to and After 2016
References
- 1.Barbieri JS, James WD, Margolis DJ. Trends in prescribing behavior of systemic agents used in the treatment of acne among dermatologists and nondermatologists: a retrospective analysis, 2004-2013. J Am Acad Dermatol. 2017;77(3):456-463. doi: 10.1016/j.jaad.2017.04.016 [DOI] [PubMed] [Google Scholar]
- 2.Barbieri JS, Choi JK, Mitra N, Margolis DJ. Frequency of treatment switching with spironolactone and oral tetracycline-class antibiotics for women with acne: a retrospective cohort study 2010-2016. J Drugs Dermatol. 2018;17(6):632-638. [PubMed] [Google Scholar]
- 3.Plovanich M, Weng QY, Mostaghimi A. Low usefulness of potassium monitoring among healthy young women taking spironolactone for acne. JAMA Dermatol. 2015;151(9):941-944. doi: 10.1001/jamadermatol.2015.34 [DOI] [PubMed] [Google Scholar]
- 4.Thiede RM, Rastogi S, Nardone B, et al. Hyperkalemia in women with acne exposed to oral spironolactone: a retrospective study from the RADAR (Research on Adverse Drug Events and Reports) program. Int J Womens Dermatol. 2019;5(3):155-157. doi: 10.1016/j.ijwd.2019.04.024 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Bird ST, Pepe SR, Etminan M, Liu X, Brophy JM, Delaney JA. The association between drospirenone and hyperkalemia: a comparative-safety study. BMC Clin Pharmacol. 2011;11:23. doi: 10.1186/1472-6904-11-23 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Krunic A, Ciurea A, Scheman A. Efficacy and tolerance of acne treatment using both spironolactone and a combined contraceptive containing drospirenone. J Am Acad Dermatol. 2008;58(1):60-62. doi: 10.1016/j.jaad.2007.09.024 [DOI] [PubMed] [Google Scholar]
- 7.Graber E. Acne vulgaris: management of moderate to severe acne. UpToDate. Published online January 14, 2020. Last updated December 9, 2020. Accessed December 22, 2020. https://www.uptodate.com
- 8.Zaenglein AL, Pathy AL, Schlosser BJ, et al. Guidelines of care for the management of acne vulgaris. J Am Acad Dermatol. 2016;74(5):945-973. doi: 10.1016/j.jaad.2015.12.037 [DOI] [PubMed] [Google Scholar]
- 9.Barbieri JS, Spaccarelli N, Margolis DJ, James WD. Approaches to limit systemic antibiotic use in acne: systemic alternatives, emerging topical therapies, dietary modification, and laser and light-based treatments. J Am Acad Dermatol. 2019;80(2):538-549. doi: 10.1016/j.jaad.2018.09.055 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Optum.com . Research data assets. Published 2015. Accessed September 5, 2018. https://www.optum.com/content/dam/optum/resources/productSheets/5302_Data_Assets_Chart_Sheet_ISPOR.pdf
- 11.von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP; STROBE Initiative . The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet. 2007;370(9596):1453-1457. doi: 10.1016/S0140-6736(07)61602-X [DOI] [PubMed] [Google Scholar]
- 12.Ejaz A, Malaiyandi V, Kim WB, Rogalska T, Alhusayen R. Validating the diagnostic code for acne in a tertiary care dermatology centre. Eur J Dermatol. 2015;25(5):469-471. doi: 10.1684/ejd.2015.2606 [DOI] [PubMed] [Google Scholar]
- 13.Barbieri JS, Weiner DM, Kakpovbia E, Nagler AR. Validating the optimal classification approach using ICD-10 codes to identify dermatology patients with acne. J Am Acad Dermatol. Published online July 22, 2020. doi: 10.1016/j.jaad.2020.07.075 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Anderson KL, Dothard EH, Huang KE, Feldman SR. Frequency of primary nonadherence to acne treatment. JAMA Dermatol. 2015;151(6):623-626. doi: 10.1001/jamadermatol.2014.5254 [DOI] [PubMed] [Google Scholar]
- 15.de Lucas R, Moreno-Arias G, Perez-López M, Vera-Casaño Á, Aladren S, Milani M; ACTUO Investigators study group . Adherence to drug treatments and adjuvant barrier repair therapies are key factors for clinical improvement in mild to moderate acne: the ACTUO observational prospective multicenter cohort trial in 643 patients. BMC Dermatol. 2015;15:17. doi: 10.1186/s12895-015-0036-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Nieuwlaat R, Wilczynski N, Navarro T, et al. Interventions for enhancing medication adherence. Cochrane Database Syst Rev. 2014;11(11):CD000011. doi: 10.1002/14651858.CD000011.pub4 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Feldman SR, Camacho FT, Krejci-Manwaring J, Carroll CL, Balkrishnan R. Adherence to topical therapy increases around the time of office visits. J Am Acad Dermatol. 2007;57(1):81-83. doi: 10.1016/j.jaad.2007.04.005 [DOI] [PubMed] [Google Scholar]
- 18.Straight CE, Lee YH, Liu G, Kirby JS. Duration of oral antibiotic therapy for the treatment of adult acne: a retrospective analysis investigating adherence to guideline recommendations and opportunities for cost-savings. J Am Acad Dermatol. 2015;72(5):822-827. doi: 10.1016/j.jaad.2015.01.048 [DOI] [PubMed] [Google Scholar]
- 19.Lee YH, Liu G, Thiboutot DM, Leslie DL, Kirby JS. A retrospective analysis of the duration of oral antibiotic therapy for the treatment of acne among adolescents: investigating practice gaps and potential cost-savings. J Am Acad Dermatol. 2014;71(1):70-76. doi: 10.1016/j.jaad.2014.02.031 [DOI] [PubMed] [Google Scholar]
- 20.Barbieri JS, Hoffstad O, Margolis DJ. Duration of oral tetracycline-class antibiotic therapy and use of topical retinoids for the treatment of acne among general practitioners (GP): a retrospective cohort study. J Am Acad Dermatol. 2016;75(6):1142-1150. doi: 10.1016/j.jaad.2016.06.057 [DOI] [PubMed] [Google Scholar]
- 21.Morris ZS, Wooding S, Grant J. The answer is 17 years, what is the question: understanding time lags in translational research. J R Soc Med. 2011;104(12):510-520. doi: 10.1258/jrsm.2011.110180 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Gupta DM, Boland RJ Jr, Aron DC. The physician’s experience of changing clinical practice: a struggle to unlearn. Implement Sci. 2017;12(1):28. doi: 10.1186/s13012-017-0555-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Guzman AK, Barbieri JS. Comparative analysis of prescribing patterns of tetracycline class antibiotics and spironolactone between advanced practice providers and physicians in the treatment of acne vulgaris. J Am Acad Dermatol. Published online June 16, 2020. doi: 10.1016/j.jaad.2020.06.044 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eFigure 1. Temporal Trends in Frequency of Baseline Potassium Labs Among Women Treated with Spironolactone for Acne
eFigure 2. Temporal Trends in Frequency of Monitoring Potassium Labs Among Women Treated with Spironolactone for Acne
eFigure 3. Variation in Baseline and Routine Potassium Monitoring Among Dermatologists Prior to and After 2016
eFigure 4. Variation in Baseline and Routine Potassium Monitoring Among Internists Prior to and After 2016
