2006 Limited FDA Survey of Compounded Drug Products

Executive Summary

Several studies, including a survey conducted by FDA in 2001, have reported quality problems with various pharmacy-compounded drugs, including sub-potency, super-potency, and contamination.  To explore these quality issues, FDA conducted an additional survey of compounded drug products in 2006.  FDA collected both active pharmaceutical ingredient (API) and finished compounded drug product samples during unannounced visits to compounding pharmacies located throughout the country.  The samples were sent to FDA field laboratories for chemical analysis to measure identity of active ingredients, potency, and uniformity of dosage.  The analytical methods used were generally United States Pharmacopoeia (USP) or modified‑USP methods.  Once all analyses were complete, FDA staff evaluated the analytical data and methods corresponding to all samples that failed at least one analytical test. 

Among 198 samples gathered during FY2006, 125 were APIs and 73 were compounded finished drug products.  The samples comprised three major drug classes:  female hormone products, inhalation products, and local anesthetic products.    

All 125 API samples passed analysis.  Of the 73 compounded finished drug products, sixteen samples were not analyzed because the expiration dates on the samples elapsed before analysis.  The remaining 57 samples were analyzed, but the results of the analyses for 21 of these samples were deemed unusable for various reasons and excluded from the survey. Of the remaining 36 samples, 12 (33%) failed analytical testing using rigorously defensible testing methodology. 

Most of the products that failed analysis did so due to sub or super-potency, called assay, or a lack of uniformity of individual dosage units, called content uniformity.  Potency ranged from 67.5% to 268.4% of the amount of drug declared on the product labeling.  For content uniformity analysis of products containing multiple active components, both sub- and super-potent active components were found within the same product samples.  Such variability can lead to uncertainty in dosing and raises concern for patient therapy.

The results of the survey suggest that problems with the quality of compounded drugs occur throughout the country.  Of note is that all APIs passed analytical testing, supporting the notion that the observed failures of the finished drug products may be causally related to the compounding processes at pharmacies. 

Introduction

FDA regards traditional pharmacy compounding as the extemporaneous combining, mixing, or altering of ingredients by a pharmacist in response to a physician’s prescription to create a medication tailored to the specialized medical needs of an individual patient. Traditional compounding typically occurs when an FDA-approved drug is unavailable or a licensed health‑care provider decides that an FDA-approved drug is not appropriate for his or her patient’s medical needs.  By definition, pharmacy compounding involves making a new drug for which safety and efficacy have not been demonstrated with the kind of data that FDA requires to approve a new drug.  In virtually all cases, FDA regards compounded drugs as unapproved new drugs.

The unapproved status of compounded drugs notwithstanding, FDA has long recognized that traditional pharmacy compounding serves an important public health function.  FDA has historically exercised enforcement discretion and generally has not taken enforcement action against pharmacies engaged in traditional compounding.  Rather, FDA has directed its enforcement resources toward firms that manufacture large quantities of unapproved new drugs under the guise of traditional compounding, and whose compounding practices result in significant violations of the new drug, adulteration, or misbranding provisions of the FDCA.

Improperly compounded, adulterated drugs have the potential to cause significant harm to patients.  From 1990 to 2005, FDA learned of at least 240 serious illnesses and deaths associated with improperly compounded products.  Because pharmacists are not required to report adverse events to FDA, there may be additional deaths and injuries of which the agency is unaware.  There are also reports in the literature of serious injuries, some involving children, caused by improperly compounded products.  For instance, two cases of hospitalization of children occurred due to compounded preparations that were 10-fold and 87-fold superpotent.¹  In another case, pharmacy mixing of a medicine too long before use, allowing chemical generation of a toxin, was suggested to have a role in the death of a patient.²  Finally, a retrospective review of patient records at a poison control center for a 35 month period found three overdoses, two involving medications for children, related to compounded drugs.³   These cases illustrate that the quality problems associated with compounded drug products can have serious and fatal consequences for patients. 

FDA is aware of a number of product quality problems associated with compounded drugs including contamination, poor compounding processes, and product toxicity.⁴  In 2001, FDA conducted limited sampling of compounded drug products purchased over the Internet.⁵  This preliminary study showed that 9 of 29 samples analyzed (31%) were sub-potent, ranging from 59% to 89% of the labeled value. Issues with product potency have been observed in other studies and surveys.  For example, Azarnoff et al.⁶ filled prescriptions for 0.3% nitroglycerin ointment and, upon analysis, found that 29% of the 24 samples were subpotent and one sample was superpotent.  This product was widely used, as it was reported that 84,000 prescriptions for this ointment were written in 2004.  In another survey, compounded drug testing conducted by the Missouri State Board of Pharmacy demonstrated failure rates ranging from 19.8% to 25.2% of all compounded drugs tested, including samples that contained 0.0% to 553% of the product’s labeled potency.⁷ 

The cause of errors in compounding leading to incorrect potency is not well understood, but such compounding errors can be observed during pharmacy student training.  Kadi and colleagues⁸ examined the accuracy of the compounding of two simple solutions by pharmacy students.  For one of the solutions, only 54% of students prepared the medications within 10% of the desired concentration.  Errors for the remaining mixtures ranged from less than 75% to greater than 200% of the desired concentration.  Results for the second solution showed 78% of students within +/- 10% of the desired concentration, however, the range of concentration errors was greater (-89% to 269%).  

In addition to incorrect potency, contamination of compounded finished drug products has also been observed.  Contamination can manifest as either microbial contamination or the presence of chemical impurities.  Several studies have shown that both incorrect potency and contamination can exist within the same products.  Goldman⁹ conducted an analysis of solutions of sodium tetradecyl sulfate from three compounding pharmacies and found that the dosage forms failed to meet labeled concentrations.  The analysis also detected the presence of the contaminant carbitol in each sample.  In another study, Mahaguna and colleagues¹⁰ reported the analysis of compounded progesterone suppositories from 10 randomly selected pharmacies throughout the United States.  According to the study, 9 of 10 pharmacies provided suppositories that fell outside potency limits set for approved products.  In addition, one pharmacy provided suppositories testing positive for Comamonas acidovorans bacteria.  Another study, by Sarver et al,¹¹ found that 18% of extemporaneously compounded samples of alprostadil for injection tested positive for microbial contamination.  In addition, 23% of samples had alprostadil concentrations of greater than +/- 10% of their labeled amount.  

The quality problems encountered with compounded finished drug products are magnified by their extensive use.  In a survey of independent community pharmacies in Illinois, Missouri, Kansas, and Iowa, 94% of pharmacies reported that 2.3% of their dispensed prescriptions were compounded.¹²  On a nationwide basis, one percent of all prescriptions, totaling approximately 30 million, were estimated to be compounded in 2003.¹³  Given the widespread occurrence of pharmacy compounding and the potential for serious injury resulting from the use of improperly compounded products, the quality of compounded drugs creates an important public health concern.

Methods

The present survey was designed collaboratively by FDA’s Center for Drug Evaluation and Research, Office of Compliance, and FDA’s Office of Regulatory Affairs.  FDA staff compiled a list of compounded drugs that are both commonly used and likely to be available at compounding pharmacies, in addition to compounded drugs that could be associated with quality or compounding problems.  Investigators obtained samples of both APIs and finished compounded drugs from compounding pharmacies located throughout the country.  Most pharmacies were chosen from a list of compounding pharmacies that was created from the internet and telephone directories.  The samples were sent to FDA field laboratories for analysis. 

The samples were analyzed for identification of active ingredient as well as potency, or assay, of ingredients declared in the product labeling.  If the compounded finished product was in capsule form, the capsules were analyzed for content uniformity.  The analytical methods were from the United States Pharmacopeia (USP).  High pressure liquid chromatography was generally used to quantify the amount of active drug substances present (titration was used once).  Failure of a sample to meet USP specifications for the drug as labeled was confirmed with an additional analysis, known as a “check analysis,” by a second chemist.  Any analyses that occurred past the labeled date of expiration were considered invalid.  The collection reports and laboratory analysis results were entered into FDA’s database of field activities:  the Field Accomplishments and Compliance Tracking System (FACTS). 

Documentation in FACTS was retrieved for summarization.  All analytical worksheet documentation for drug samples failing USP specifications was collected and scrutinized.  Based on the quality of the analytical data, the laboratory analysis for each sample that failed analytical testing was classified as acceptable (i.e., scientifically rigorous) or unacceptable.

Back to Top

Results

FDA collected 125 API samples and 73 compounded finished product samples from pharmacies located throughout the United States (Table 1).  The samples encompassed three major therapeutic areas:  local anesthetic products, inhalation products, and female hormone products.  Both the finished compounded drug products and the APIs were analyzed as described in the Methods section.  Table 2 summarizes the results of the laboratory analyses.  All 125 of the APIs passed laboratory analysis for both the identity of the active drug substance and assay for the amount of active drug substance present. 

Of the 73 compounded finished drug products, sixteen samples were not analyzed because the expiration dates on the samples elapsed before analysis.  The remaining 57 samples were analyzed, but the results of the analyses for 21 of these samples were deemed unusable for various reasons and excluded from the survey. Of the remaining 36 samples, 12 (33%) failed analytical testing using rigorously defensible testing methodology (Table 2). 

The compounded finished drug products that were analyzed included different formulations and fell into three therapeutic classes:  female hormones, local anesthetics, and inhalation drugs.  Analysis results for these products, categorized by both formulation and therapeutic class, are described in Table 3.  Three local anesthetic products failed analysis.  For female hormone products, 9/31 (29%) failed analysis.  Two inhalation products were tested and passed analysis.   

The failed samples passed identity testing but failed assay and, in some cases, content uniformity testing.  Content uniformity testing was only performed on compounded finished products in capsule form.  Table 4 summarizes the number of products for each formulation that failed testing and whether the samples failed assay testing, content uniformity testing, or both.  All products that failed at least one analytical test failed assay testing, suggesting inadequate potency, and the vast majority of capsules failed content uniformity testing.  Potency ranged from 67.5% to 268.4% of the amount of drug declared on the product labeling.

Capsules with content uniformity problems often exhibited potency variations with respect to more than one active component.  It was possible to find both sub-potent and super-potent active components within the same sample.  Figure 1 illustrates the content uniformity observed in a sample in which capsules appear to contain super-potent concentrations of estrone and sub-potent concentrations of estriol and estradiol. 
  

Discussion

These data reflect significant problems with the quality of certain compounded drugs.  The analysis of compounded finished drugs revealed instances of unreliable potency and, in the case of products in capsule form, unreliable uniformity of dosage.  The observed potencies of the finished drug products were widely variable, ranging from 67.5% to 268.4% of the labeled potency.  For capsules, both sub- and super-potent active components were found in the same product samples.  Variability in potency can lead to uncertainty in dosing and raises concerns for patient therapy.

All of the API samples passed analytical testing, suggesting that the failures of the compounded finished drug samples were related to compounding processes at the pharmacies.  Previous studies suggest that seemingly straightforward preparation of certain drugs may present significant challenges in pharmacy settings,10 particularly when proper in-process controls and end-product testing are not in place to monitor and ensure dosage quality and uniformity.13  For instance, the uniformity of multi-component solid oral dosage forms relies on achieving homogeneity of a powder mixture prior to encapsulation.  This process of mixing a multi-component product is scientifically complex and may be influenced by numerous factors including, but not limited to, the physical and chemical properties of the ingredients and their interactions with each other, properties of the mixing container, speed of mixing, temperature, and humidity.^{14, 15, 16, 17, 18} 

Some samples were excluded from the study for procedural and methodological reasons.  These include sixteen products that were not analyzed because they reached their labeled expiration date prior to analysis.  In addition, 21 samples failed analysis, but were excluded due to one or more of the following methodology problems:  the samples had expired when tested, the expiration date passed before check analysis could be performed, the analytical method was technically poor, or the sample passed the check analysis after failing initial testing (Table 5). 

The limitations of this survey included resource constraints and issues related to laboratory analysis.  During laboratory analysis, there were instances when methodology development and validation were needed to accurately analyze the samples.  Because the samples were collected in limited quantities, this was not always possible.  In those instances, the test results were excluded from the survey.  In addition, potency, identity, and content uniformity were each assessed with a single analytical technique.  The fact that no secondary analytical methods were used for the sample analyses reflects the informational nature of the survey.

Back to Top

Conclusions

Poor quality compounded drugs are a serious public health concern, as improperly compounded products have been linked to grave adverse events, including deaths.  The results of the FY 2006 compounded drug survey suggest a problem with the quality of certain compounded drug products.  All of the API samples passed analytical testing, suggesting that the analytical failures of the finished drug products were likely related to the compounding processes at the pharmacies. Thirty-three percent of compounded finished product samples did not conform to product labeling in terms of potency and/or content uniformity.  The majority of the finished compounded product samples analyzed in this survey were hormone therapy products.  These products appear to be popular, commonly compounded, and available at many pharmacies.  Given the wide use of these products, the fact that nearly one-third failed analytical testing raises public health concerns.  The potencies of the compounded finished drug products that failed analysis were highly variable and, in the case of content uniformity, both sub- and super-potent active components were found in the same product samples.  The wide range in potency and uniformity of dosage observed in this survey can, in a clinical setting, lead to uncertainty in the amount of drug that a patient receives.  Such uncertainty can lead to medication errors that pose a health risk for patients who rely on compounded drugs.

Table 1.  Active Pharmaceutical Ingredients and Finished Compounded Drug Products Collected for the FY 2006 Compounded Drugs Survey

Back to Top

Table 2.  Results of Laboratory Analysis for all Survey Samples

* Samples excluded from survey

Table 3.  Analysis Results of Compounded Finished Drug Products Categorized by Formulation and Therapeutic Class

Table 4.  Compounded Finished Products that Failed Assay for Potency or Content Uniformity

* Four products failed initial testing, but check analyses were not performed, samples excluded

 Figure 1.  Lack of Uniformity of Dosage for Capsules Containing Multiple Active Components

Table 5.  Characterization of Methodology Issues Encountered During FY 2006 Compounded Drugs Survey (Samples could meet more than one criterion.)

Acknowledgments
Contributors to the study include individuals from numerous FDA units described below:

Survey design:  Compounded Drugs Team, Division of New Drugs & Labeling Compliance, Office of Compliance, Center for Drug Evaluation and Research (CDER)
Survey assignment development:  Surveillance Programs Team (SPT), Division of Compliance Risk Management and Surveillance, Office of Compliance, CDER
Sample collection:  15 Districts in Office of Regulatory Affairs (ORA) –

• Atlanta
• Baltimore
• Chicago
• Cincinnati
• Dallas
• Detroit
• Florida
• Los Angeles
• Minneapolis
• New England
• New Jersey
• New York
• Philadelphia
• San Francisco
• San Juan

Laboratory assignments and laboratory worksheet review:  Division of Field Science, ORA
Laboratory analysis:  9 Laboratories in ORA –

• Denver District
• Detroit District
• Kansas City District
• Northeast Regional
• Philadelphia District
• Pacific Regional – Northwest (Seattle)
• Pacific Regional – Southwest (Los Angeles)
• San Juan District
• Southeast Regional

Data assembly, analysis, interpretation, and report:  SPT.

¹ Suchard JR.  Graeme KA.  Pediatric clonidine poisoning as a result of pharmacy compounding error.  Pediatric Emergency Care.  2002; 18(4):295-296.
² FDA Alert: Colistimethate (marketed as Coly-Mycin M and generic products) Information.  US Food and Drug Administration.  June 28, 2007.  http://www.fda.gov/cder/drug/infopage/colistimethate/default.htm.
³ Seifert SS, Jacobitz K.  Pharmacy prescription dispensing errors reported to a regional poison control center.  J Toxicology Clinical Toxicology 2002; 40(7):919-923. 
⁴ Consumer Update: The special risks of pharmacy compounding.  Food and Drug Administration.  May 31, 2007.  Http://www.fda.gov/consumer/updates/compounding 053107.html.
⁵ Report: Limited FDA Survey of Compounded Drug Products.  Food and Drug Administration.  January 28, 2003.  http://www.fda.gov/cder/pharmcomp/survey.htm.
⁶ Azarnoff DL, Lee JC, Lee C, Chandler J, Karlin D.  Quality of extemporaneously compounded nitroglycerin ointment.  Dis Colon Rectum 2007 Apr; 50(4):509-16.
⁷ Compounded Drug Testing Report Fiscal Year 2006.  Missouri State Board of Pharmacy.  http://pr.mo.gov/boards/pharmacy/Pharmacy-Compounding-Report-FY-2006.pdf.
⁸ Kadi A, Francioni-Proffitt D, Hindle M, Soine W.  Evaluation of basic compounding skills of pharmacy students.  Am J Pharmaceutical Education 2005; 69 (4):508-515.
⁹ Goldman MP.  Sodium tetradecyl sulfate for sclerotherapy treatment of veins: is compounding pharmacy solution safe?  Dermatol Surg 2004; 30(12): 1454-1456.
¹⁰ Mahaguna V, McDermott JM, Zhang F, Ochoa F.  Investigation of product quality between extemporaneously compounded progesterone vaginal suppositories and an approved progesterone vaginal gel.  Drug Dev Ind Pharm 2004; 30(10):1069-78.
¹¹ Sarver JG, Peng N, Lerdkanchanaporn S, Oravecz-Wilson K, Alexander KS, Erhardt PW.  Analysis of extemporaneous alprostadil formulations.  Int J Pharmaceutical Compounding.  1999 Mar Apr;3(2):148-155.
¹² McPherson TB, Fontane PE, Jackson KD, Martin KS, Berry T, Chereson R, Bilger R.  Prevalence of compounding in independent community pharmacy practice.  J Am Pharm Assoc (2003).  2006 Sep-Oct; 46(5):568-73.
¹³ Statement by Steven K. Galson, M.D., M.P.H., Acting Director, Center For Drug Evaluation and Research,  
U.S. Food and Drug Administration, Department of Health And Human Services, before the Senate Committee on Health, Education, Labor, and Pensions hearing on “Federal And State Role In Pharmacy Compounding And Reconstitution: Exploring The Right Mix To Protect Patients.”  October 23, 2003.  Food and Drug Administration.  http://www.fda.gov/ola/2003/pharmacycompound1023.html.
¹⁴ Lachman L, Lieberman HA, Kanig JL.  Theory and practice of industrial pharmacy.  3rd ed.  Lea & Febiger, Philadelphia, 1986.
¹⁵ Gennaro AR, ed.  Remington’s Pharmaceutical sciences.  18th ed.  Mack Publishing Co., Easton, PA, 1990.
¹⁶ Nash RA, Wachter AH.  Lachman.  Theory and practice of industrial pharmacy.  Marcel Dekker, Inc., New York, 2003.
¹⁷ Willig SH, Stoker JR.  Good manufacturing practices for pharmaceuticals: a plan for total quality control.  3rd ed.  Marcel Dekker, Inc., New York, 1992.
¹⁸ Sharp J.  Good manufacturing practice: philosophy and applications.  Interpharm Press, Buffalo Grove, IL, 1991.

Back to Top     Back to Pharmacy Compounding

PDF requires the free Adobe Acrobat Reader

Date created: January 28, 2003, Updated March 5, 2008

CDER Home Page | CDER Site Info | Contact CDER | What's New @ CDER
FDA Home Page | Search FDA Site | FDA A-Z Index | Contact FDA | Privacy | Accessibility | HHS Home Page

FDA/Center for Drug Evaluation and Research