Gender Bias in Studies for Food and Drug Administration Premarket Approval of Cardiovascular Devices
Background—Cardiovascular devices can have different safety and effectiveness profiles in men and women. The type and quality of sex-specific data reviewed by the Food and Drug Administration (FDA) before approval of these devices are unknown.
Methods and Results—We performed a systematic review of the demographics, comments on gender bias, and analysis of results by sex for 78 high-risk cardiovascular devices that received premarket approval by the FDA between 2000 and 2007. FDA summaries of evidence did not report sex of enrollees in 34 (28%) of 123 studies. For studies reporting sex distribution, the study populations were, on average, 67% men. There was no increase in the enrollment of women over time. Explanations for the relatively low percentage of women often stated that the trials reflected either underlying disease distribution or referral rates for similar procedures or that the sex distribution reflected similar or previous trials. Forty-one percent of studies included a gender bias comment or analysis, and 12 (26%) of 47 of these analyses identified some difference in device safety or effectiveness by sex.
Conclusions—There is a lack of sex-specific safety and effectiveness data for high-risk cardiovascular devices before FDA approval. The justifications for this lack of evidence may perpetuate the status quo. More rigorous FDA requirements for sex-specific data before device approval could present an opportunity to improve cardiovascular outcomes.
In 1985, the US Public Health Service Task Force concluded that “the historical lack of research focus on women's health concerns has compromised the quality of health information available to women as well as the health care they receive.”1 The following year, the National Institutes of Health (NIH) adopted a policy statement encouraging inclusion of women and minorities in all federally funded research,2 and the Food and Drug Administration (FDA) in 1988 specifically called for studies of whether safety and effectiveness were similar within population subgroups defined by characteristics such as age, sex, and race.3
Around this time, the medical research community was gaining appreciation for the plausible biological basis of sex differences in disease. In 1990, the American Medical Association Council on Ethical and Judicial Affairs concluded that “medical treatments for women are based on a male model, regardless of the fact that women may react differently to treatments than men or that some diseases manifest themselves differently.”4 In response to concerns raised by medical researchers and women's health advocates, the Government Accountability Office audited clinical trials funded by the NIH or submitted to the FDA in support of drug marketing applications and in a 1990 report, observed “little progress” by the NIH in implementing its inclusion policies and concluded that these had not been well communicated or understood within NIH or the research community.5 Consequently, the NIH created the Office of Research on Women's Health and required inclusion of women and minorities, unless there was a clear justification for their exclusion, as well as analysis of results by sex in phase III trials. Congress strengthened the authority of these guidelines with the NIH Revitalization Act of 1993.6 A follow-up Government Accountability Office audit in 2000 concluded that NIH-funded trials include women proportionate to their numbers in the general population, but valid analysis of results by sex was still wanting, and there was noted to be scant evidence that NIH staff and reviewers were implementing the requirement.7
In 1994, the Office of Women's Health was established within the FDA. That year, the Center for Devices and Radiological Health instituted a policy to address the possibility of gender bias in submissions and review documentation for new medical devices.8 Gender bias is a term the FDA uses in its 1994 directive and is the specific language found in the FDA Summaries of Safety and Effectiveness Data (SSEDs). Specifically, this directive stated:
Gender bias needs to be addressed in writing in the SSED from two aspects for all pending as well as future PMA [premarket approval] submissions.
Was the selection ratio of men versus women in the study reflective of the underlying distribution of the disease for that given age group, ethnic group, stage of disease, etc.? Was any selection bias on the basis of gender identified during review?
Was there any difference in the safety and effectiveness of the device based on gender? For example, was the device more/less effective in women?8
There have been no analyses of gender bias in studies submitted for FDA approval of medical devices. We focused on cardiovascular devices given their impact on morbidity and mortality. Many factors, such as differences in body size and bleeding tendencies, could contribute to sex differences in device safety profiles and outcomes.9 Therefore, it is important to have sex-specific cardiovascular device data. The 2001 Institute of Medicine report “Does Sex Matter?” recommends using the term gender for cultural issues and the term sex for biological issues. However, we use gender bias to be consistent with FDA terminology.
The review process for medical devices is determined by their level of risk to patients.10 There are 3 classes of devices: Class I devices (eg, elastic bandages and examination gloves) are considered low risk, class II devices (eg, powered wheelchairs and infusion pumps) are moderate risk, and class III devices (eg, heart valves and implanted cerebella stimulators) are high risk.10,11 Class III devices “are usually those that support or sustain human life, are of substantial importance in preventing impairment of human health, or which present a potential, unreasonable risk of illness or injury.” The regulatory process for class III devices is generally through PMA. According to the FDA, this review process is “the most stringent type of device marketing application required.”12
The safety and effectiveness of cardiovascular devices may differ by sex.
Several national initiatives and guidelines over the past 2 decades have attempted to increase the amount of available sex-specific data for drugs and devices.
Despite Food and Drug Administration policy to analyze gender bias in all device applications, the majority of high-risk cardiovascular devices approved by the administration still do not contain such information.
Policy changes are necessary to ensure analysis of sex-specific data during device evaluation in order to optimize safety and effectiveness data.
After a device receives premarket approval (PMA), the FDA makes publicly available its approval order, labeling (instructions for use), and an SSED. The SSED “is intended to present a reasoned, objective, and balanced critique of the scientific evidence which served as the basis of the decision to approve or deny the PMA.”13 SSEDs are where gender bias comments or gender analyses instituted by the Center for Devices and Radiological Health in 1994 can be found. Given that the clinical evidence for class III devices undergoes the most rigorous evaluation by the FDA, we chose to study gender bias comments and analyses in class III devices. We hypothesized that sex-specific reporting and gender bias comments would improve over time for these cardiovascular devices.
We conducted a systematic review of all gender bias comments, reporting of demographics with regard to sex (percentage of men versus women), and analyses of results for all approved PMA applications for high-risk cardiovascular devices submitted to the FDA from 2000 through 2007 in order to better understand the extent of sex-specific data available for and used by the FDA in the evaluation of cardiovascular devices. Our study is limited to approved PMA applications because the FDA does not make available SSEDs for devices that were not approved. To analyze gender bias comments, we specifically looked for sections in the SSEDs that stated “Gender Bias Comment” or “Gender Analysis.” We examined these sections for the characteristics described previously.
Our review was conducted using SSEDs available at the time of a search performed on October 15, 2008; this time frame was chosen to enable a thorough compilation of PMAs during this decade and allow time for SSEDs of the most recent devices reviewed by the FDA to be made publicly available.
Data Acquisition and Coding of Enrolled Patients
The initial data collection methods for 78 PMAs and the number of enrolled patients in each of the studies as well as the demographic data have been described previously.14 When the percentage of men was not available in the SSEDs, if the device had been sent to the Circulatory System Devices Panel, we reviewed the given panel meeting notes for these data. In addition, all data related to sex proportions and justifications for sex proportions and sex-specific data were collected and coded for each study (Figure). Textual descriptions in the SSEDs of all gender bias comments also were examined.
For each category for which data were coded, data were tabulated across studies. These summary data are presented as number of studies. Studies with fewer than 50 enrolled patients or where the highest number of patients mentioned in the SSED was fewer than 50 were excluded from all calculations under the subset of “for studies with a gender bias comment or gender analysis” or, if there was no gender bias comment or analysis, from the examination of data or statistical analyses stratified by sex. Devices also were classified into the following categories for analyses: cardiac stents, noncardiac stents, bridge to transplant, intracardiac devices, hemostasis devices, electrophysiology devices, endovascular grafts, and miscellaneous/other. For calculations summarizing data for primary end points, our analyses excluded those studies in which we were unable to determine any primary end point.
Associations between study-level covariates and the proportion of women (gender bias) were estimated by random effects meta-regression, using the Stata “metareg” function with restricted maximum likelihood estimation. The dependent variable was the proportion of women, and independent variables were potential predictors of gender bias. Potential predictors considered separately were date received, mean age, and device category. To account for correlation between standardized effects sizes within a device, each meta-regression model was bootstrapped (2000 replications) at the level of device, and 95% CIs and P values based on bootstrap were computed. To examine whether the presence of comments on gender bias, comments on sex proportions reflecting previous trials, or comments of sex distribution or referral changed with time, random effects logistic regression was performed with a random effect for device. The independent variable was date received. All analyses were performed using Stata version 10 statistical software.
Enrollment by Sex
We examined data on enrollment by sex in all 123 studies from 78 PMAs from 2000 to 2007 (online-only Data Supplement Table). These data were in the SSED for 89 (72%) studies and were located in Circulatory System Device Panel meeting data for an additional 4 (3%) studies after review of 5 meetings for PMAs of studies in which percentage of men was not available in the SSED. These 93 studies had a mean of 67% men (range, 28% to 92%). Eighty-two (88%) studies had more men enrolled than women. Results of the meta-regression indicate that there was no increase in percentage of women over the 8-year period (P=0.957). Mean age was a statistically significant predictor of sex, with each increase in mean age of 10 years associated with a decrease in 5% in the percentage of women (P<0.0005).
We used electrophysiology devices (72% men) as the reference because this category was the largest. We found a significant difference in percentage of men only compared with intracardiac devices, such as heart valves and atrial and ventricular septal occluders (52% men; P=0.001).
In 80 studies, both the number enrolled and the number for the percentage of men were identified. Of these, 38 (48%) had a discrepancy between those values (Table 1). Thirty-two (84%) of these discrepancies were due to more enrolled patients than for whom the percentage of men was provided; for these studies, an average of 40 patients were enrolled, but their sex proportion was not provided (range, 1 to 296). In 6 (16%) of 38 studies, there was a greater number of patients for whom the percentage of men was provided than the number of patients enrolled; all of these cases were due to use of retrospective controls. In this case, the mean discrepancy was 235 (range, 49 to 588) excess analyzed patients.
Gender Bias Comments or Analyses
Of 123 studies, 51 (41%) had a gender bias comment or analysis (Table 2). Only 1 comment referred to more women enrolled than men. One study was excluded from further analysis because it enrolled fewer than 50 patients. Forty-seven (94%) of the 50 comments discussed an examination of any study results by sex. About one fourth (12/47) of these studies described some difference in device safety or effectiveness by sex.
Raw data on outcome stratification by sex was provided in 19 (38%) of the 50 studies with gender bias comments or analysis, and 10 (53%) of these 19 referred to a primary end point. Statistical data on outcome stratification by sex was provided in 8 (16%) of the 50 studies with gender bias comments or analysis. Of these 50 studies, 5 (10%) referred specifically to a primary end point.
In 57 studies without a gender bias comment, sex-specific data or analysis reporting occurred in 6 (11%). Fifteen (12%) of 123 studies had fewer than 50 patients enrolled and, thus, were excluded from this analysis.
Studies with a gender bias comment or analysis did not differ in the percentage of women included compared to studies without such comments or analyses (P=0.487). There was no change in the presence of gender bias comments or analyses over the 8-year period that PMAs were received by the FDA (P=0.490).
Statements Regarding Study Sex Proportions
In 36 (29%) of 123 studies, the study sex proportions were declared to reflect the disease distribution or referral for that or similar procedures (Table 2). These statements were sometimes inaccurate. One example is a trial enrolling 86% men that stated, “The occurrence of AAA [abdominal aortic aneurysm] disease is known to be higher in men than women and the ratio of men to women enrolled in this study reflects the general population,”15 and references an article by Lee et al.16 However, this citation does not support the statement in the SSED. It refers to a 47-patient surgical study in which 41 (87%) were men and is not a study of the sex distribution of patients with AAA. In studies that do look at demographics, women comprise 22% to 25% of patients with AAAs. 17,18 Another example, for an implantable cardioverter-defibrillator (ICD), stated, “There were 15 females out of 126 patients (or 11.9%). The relatively low percentage of females enrolled into the study is related to the incidence of heart disease. If females were just as likely to have heart disease as males, then you would expect the percentage to be much closer to 50%.”19 This statement in the SSED inaccurately implies that there is a very low percentage of women with heart failure and that is why the trial enrolled few women. For each increase in PMA receipt date by 1 year, the odds of such a statement on sex distribution decreased by 65% (P=0.020).
Of 123 studies, the low proportion of women in 8 (7%) were justified as being similar to previous trials (Table 2). One example is this statement regarding a heart failure study of which 22% were women stated, “The demographics of the study population are typical for a CRT-D [Cardiac Resynchronization Therapy Defibrillator] study performed in the US.”20 However, a study examining a national sample of Medicare beneficiaries indicated that 49% of patients with impaired left ventricular systolic function are women.21 Another SSED states:
No gender bias was noted in The Driver™ Registry. Of the total patients enrolled, 68.1% (203/298) were male. This compares quite closely with the percentage of male patients enrolled in a competitive trial, (GDT Vision, where 68.2% (182/267) [sic] of the patient population were male. These percentages reflect the percentage of males in the US cardiac population as a whole.22
There was no change in the number of such comments over the 8-year period studied in which PMA applications were received by the FDA (P=0.592).
Statements About Selection Bias
Of 123 studies, 22 (18%) had a specific statement either that there was a random selection of men and women or that no selection bias was present for either sex (Table 2). An example states, “The gender selection in this clinical trial was completely random, and patient selection was solely based on exclusion and inclusion criteria. Men represented 65% of the population.”23 There was no relationship between the presence of such a statement and the sex proportion in the study (P=0.866).
Our analysis of SSEDs for cardiovascular PMAs submitted and approved from 2000 through 2007 shows that the average proportion of women enrolled remained unchanged at approximately one third of the total study population. Importantly, we noted a lack of reported data regarding women in these device trials: 28% (34/123) of studies do not have sex proportions reported in SSEDs. Of studies that do report sex proportions, nearly one half do not report the proportions for all the patients that they enrolled, which suggests an important source of missing data. Women must be included in clinical trials, and sex-specific results must be provided for analysis. However, just 38% of studies with a gender bias comment or analysis provided such data, and even fewer (11%) studies without such a comment did so.
Although gender bias comments or analyses are required by the FDA for all studies, they were present for only 41% of studies listed in the SSEDs. Raw data or statistical analyses rarely were provided, and sex-specific analyses of study primary end points were even less common.
Some statements in these SSEDs accept the status quo as the reason for low numbers of women enrolled. Some comments justifying the low percentage of women in clinical trials link the lack of adequate numbers of women to a low referral rate for device-related procedures. This lower referral rate may be due to less use of specialist consultation for women with coronary artery disease and congestive heart failure than for men.24 The continued underrepresentation of women in cardiovascular device trials and the resultant lack of sex-specific data may perpetuate the status quo. Indeed, for National Heart, Lung, and Blood Institute-sponsored phase III and IV trials of coronary artery disease, congestive heart failure, sudden cardiac death, atrial fibrillation, hypertension, and cardiovascular disease, there is a higher disease burden in women than their representation in clinical trials.25 More precise and rigorous standards for the reporting of sex-specific data for FDA-approved devices would provide a significant improvement in the amount and quality of information available for caring for women with cardiovascular disease. For example, if PMA applications missing sex-specific data were not accepted until the required information was added, it would ensure that the necessary data were included for men and women.
Few gender bias comments or analyses provided raw data, and even fewer provided statistical analyses in the SSEDs, causing concern that unsubstantiated results may not be thorough and accurate. For example, for the TAXUS IV study of drug-eluting stents, the SSED, in reference to the TAXUS series of trials, including TAXUS IV, states that “no differences in safety or effectiveness were found, with respect to gender.”26 However, Lansky et al reported that in TAXUS IV, “women compared with men had higher unadjusted one-year rates of target lesion revascularization … (7.6% versus 3.2%, P=0.03),”27 which was the primary end point. Further, Table 3 in the Lansky et al article indicated that in addition to target lesion revascularization, women had a statistically significantly higher rate of target vessel revascularization (TVR) than men (10.8% versus 5.7%; P=0.03) driven by a higher rate of repeat coronary intervention (9.2% versus 4.1%; P=0.02). Although 1-year rates of TVR and repeat coronary intervention in the overall TAXUS group were statistically significantly less (ie, superior) to bare-metal stent control, this was not true for women as a subset (P=0.07 and P=0.08, respectively). One possible explanation for this discrepancy is that the primary end point in the TAXUS IV trial in the SSED is 9-month TVR, whereas in the published article, it is TVR at 1-year by sex. Therefore, it is possible that the 9-month analysis did not differ by sex and that the 12-month analysis did. The Lanksy et al article suggested that the TAXUS stent is more effective in men than in women, which is not consistent with the SSED's statement about effectiveness. Neither the SSED nor the published article performed a direct comparison of the degree of effectiveness in men versus women. Lansky et al analyzed only the women and concluded that the device is effective; however, their data actually suggests that the TAXUS stent is more effective in men than in women.
There also may be differences in claims of device efficacy and safety by sex. For example, Thoratec Corporation notes on its Web site regarding the HeartMate II Left Ventricular Assist System, “Its small size and quiet operation make the HeartMate II® suitable for a wider range of patients, including women and those of smaller stature.”28 Although this device could allow more implantation of ventricular assist devices in women, we have noted that the SSED stated that the small number of women in its PMA application made it difficult to draw conclusions regarding differences in the device's safety profile between men and women and that women had increased rates of some adverse events.29
The lack of sex-specific data before device approval may lead to delays in discovery of safety and efficacy concerns. For example, registry data of ICDs shows that women are more likely than men to experience any in-hospital adverse event and major adverse events.30 Safety concerns must be balanced by clinical benefits. However, pooled data indicate that there is no mortality benefit for ICDs for primary prevention in women with heart failure.31 We found that just 28% of participants in studies of electrophysiology devices are women. These incomplete safety and efficacy data may mean that more women are experiencing adverse events for a device lacking a mortality benefit, a situation unlikely to be ameliorated without FDA guidance.32
There are some limitations to our study. One is that we relied on the publicly available SSEDs and Circulatory System Device Panel meeting materials as our data sources. It is possible that these data are not comprehensive and that sex-specific data were reviewed and taken into consideration during device review but not included in the SSED or panel notes. However, given that the specific stated purpose of SSEDs is to justify the FDA's decision,13 the data should reflect the FDA's actual considerations. Further, SSEDs and Circulatory System Device Panel meeting notes are the only publicly available data source and, thus, provide the only available sex-specific data. Our study emphasizes the crucial need for improved access to full FDA reviews to elucidate the FDA's considerations in approving or denying a PMA application.33 Another limitation is that although we used the specific language used in SSEDs referring to gender bias, our study did not evaluate bias in the device approval process but, rather, gender-specific and related data.
Federal agencies have made several attempts to increase the enrollment of women in clinical trials and the reporting of sex-specific data. However, despite federal guidelines, there continues to be a lack of sex-specific reporting of data,34 which also has been noted in cardiovascular trials.35 The Institute of Medicine's recent report examining the impact of changes in government support for women's health research recommended that “all medical product evaluations by the Food and Drug Administration present efficacy and safety data separately for men and women.”36 In continued efforts to improve inclusion of women in cardiovascular clinical trials, the FDA held several workshops in 2008 in preparation for plans to publish a guidance document on “the study and analysis of sex/gender differences in cardiovascular medical device trials reviewed by the FDA.”37 These workshops have sought to convene all stakeholders, including clinical investigators and participants from various government agencies, industry, and patient advocacy groups, to develop practical guidelines for the design and conduct of clinical trials that provide more robust information about the safety and effectiveness of new treatments and diagnostic tools for both women and men with cardiovascular disease. Further, passage of the Heart Disease Education, Analysis and Research, and Treatment for Women Act, currently pending in Congress, would mandate sex-specific data reporting and help to address the disparities we have identified. These multipronged efforts all will help to increase enrollment of women and the reporting of sex-specific data to ensure that we can optimize risk/benefit in treatment choices for all patients.
Dr Redberg is a current member of the FDA Circulatory System Devices Panel and a member of the California Technology Assessment Forum.
We thank Charles McCulloch, PhD, and Erin Madden, MPH, for providing statistical assistance and Susan Wood, PhD, for reviewing the initial draft of this article.
The online-only Data Supplement is available at http://circoutcomes.ahajournals.org/cgi/content/full/CIRCOUTCOMES.110.958215/DC1.
- Received July 2, 2010.
- Accepted January 4, 2011.
- © 2011 American Heart Association, Inc.
Office of Research on Women's Health. Inclusion of women in research. http://orwh.od.nih.gov/inclusion.html. Accessed June 25, 2010.
Center for Drug Evaluation and Research. Guideline for the Format and Content of the Clinical and Statistical Sections of New Drug Applications. Washington, DC: Food and Drug Administration; 1988.
American Medical Association; Council on Ethical and Judicial Affairs. CEJA Report B–I-90. Gender disparities in clinical decision-making. http://www.ama-assn.org/ama1/pub/upload/mm/369/ceja_bi90.pdf. Accessed June 25, 2010.
United States General Accounting Office. National Institutes of Health: problems in implementing policy on women in study populations. http://archive.gao.gov/d48t13/141859.pdf. Accessed June 25, 2010.
National Institutes of Health Revitalization Act of 1993, Public L No. 103-43, 42 USC 289 (a)(1).
United States General Accounting Office. Women's health: NIH has increased its efforts to include women in research. http://www.gao.gov/archive/2000/he00096.pdf. Accessed June 25, 2010.
Department of Health and Human Services. Office of Device Evaluation Annual Report for Fiscal Year 1994. Washington, DC: US Department of Health and Human Services; 1994:26.
- Pardue M-L,
- Azarnoff DL,
- Berenbaum S,
- Berkley KJ,
- Fausto-Sterling A,
- Federman DD,
- Gilchrest BA,
- Grumbach MM,
- Kumanyika S,
- LaRosa JH,
- Lockshin MD,
- Panetta J,
- Sapienza C,
- Shaywitz SE,
- Vandenbergh JG,
- Willard HF
US Food and Drug Administration. Medical devices: general and special controls. http://www.fda.gov/MedicalDevices/DeviceRegulationandGuidance/Overview/GeneralandSpecialControls/default.htm. Accessed August 10, 2010.
US Food and Drug Administration. Medical devices: premarket approval (PMA). http://www.fda.gov/MedicalDevices/DeviceRegulationandGuidance/HowtoMarketYourDevice/PremarketSubmissions/PremarketApprovalPMA/default.htm. Accessed June 25, 2010.
US Food and Drug Administration. Medical Devices: PMA application contents. http://www.fda.gov/MedicalDevices/DeviceRegulationandGuidance/HowtoMarketYourDevice/PremarketSubmissions/PremarketApprovalPMA/ucm050289.htm. Accessed June 25, 2010.
Summary of Safety and Effectiveness Data. Talent™ Abdominal Stent Graft System [Premarket Approval no. P070027]. April 2008. http://www.accessdata.fda.gov/cdrh_docs/pdf7/P070027b.pdf. Accessed June 25, 2010.
- Lederle FA,
- Johnson GR,
- Wilson SE,
- Chute EP,
- Littooy FN,
- Bandyk D,
- Krupski WB,
- Barone GW,
- Acher CW,
- Ballard DJ
- Singh K,
- Bonaa KH,
- Jacobsen BK,
- Bjork L,
- Solberg S
Summary of Safety and Effectiveness. Phylax AV ICD, Model 122 382, Software Cartridge SWMI/I-GAV. 2.U, Model 128 814 [Premarket Approval no. P000009]. September 2000. http://www.accessdata.fda.gov/cdrh_docs/pdf/P000009b.pdf. Accessed June 25, 2010.
Summary of Safety and Effectiveness. Ovatio CRT-D Model 6750, Situs OTW Left Ventricular Over-the-Wire Lead Model UW28D, Situs OTW Stylet kit, Elaview 1.34 UG2 programming software [Premarket Approval no. P060027]. May 2008. http://www.accessdata.fda.gov/cdrh_docs/pdf6/P060027b.pdf. Accessed June 25, 2010.
Summary of Safety and Effectiveness Data. Driver™ Over-The-Wire, Rapid Exchange and Multi-Exchange Coronary Stent System [Premarket Approval no. P030009]. October 2003. http://www.accessdata.fda.gov/cdrh_docs/pdf3/P030009b.pdf. Accessed June 25, 2010.
Summary of Safety and Effectiveness Data. ACUITY™ Steerable Lead Models 4554, 4555, and 4556 [Premarket Approval no. P050046]. April 2007. http://www.accessdata.fda.gov/cdrh_docs/pdf5/P050046b.pdf. Accessed June 25, 2010.
- Cook NL,
- Ayanian JZ,
- Orav EJ,
- Hicks LS
- Kim ES,
- Menon V
Summary of Safety and Effectiveness Data. TAXUS™ Express2™ Paclitaxel-Eluting Coronary Stent System (Monorail and Over-the-Wire) [Premarket Approval no. P030025]. March 2004. http://www.accessdata.fda.gov/cdrh_docs/pdf3/P030025b.pdf. Accessed June 25, 2010.
- Lansky AJ,
- Costa RA,
- Mooney M,
- Midei MG,
- Lui HK,
- Strickland W,
- Mehran R,
- Leon MB,
- Russell ME,
- Ellis SG,
- Stone GW
Thoratec Corporation-Medical Professionals. HeartMate® II Left Ventricular Assist System: the new era begins. http://www.thoratec.com/medical-professionals/vad-product-information/heartmate-ll-lvad.aspx. Accessed June 25, 2010.
- Peterson PN,
- Daugherty SL,
- Wang Y,
- Vidaillet HJ,
- Heidenreich PA,
- Curtis JP,
- Masoudi FA
- Ghanbari H,
- Dalloul G,
- Hasan R,
- Daccarett M,
- Saba S,
- David S,
- Machado C
Institute of Medicine. Women's Health Research: Progress, Pitfalls, and Promise. Washington, DC: National Academies Press; 2010.