Safety of Prescribing PDE-5 Inhibitors via e-Medicine vs Traditional Medicine

Article Outline

• Abstract
• PATIENTS AND METHODS
  • Statistical Analyses
• RESULTS
• DISCUSSION
• CONCLUSION
• APPENDIX.. E-Medicine Internet Web Site Procedures
• REFERENCES
• Copyright

OBJECTIVE

To determine the safety of a US-based, state-regulated Internet system vs a multispecialty primary care system for prescribing phosphodiesterase type 5 (PDE-5) inhibitors for erectile dysfunction.

PATIENTS AND METHODS

From January 1, 2001, through December 31, 2005, 500 e-medicine clients (mean ± SD age, 47±11 years; hypertension, 60%; type 2 diabetes mellitus, 2%; mean ± SD number of medications, 0.4±0.8) vs 500 traditional medicine patients (mean ± SD age, 57±12 years; hypertension, 50%; type 2 diabetes mellitus, 23%; mean ± SD number of medications, 5.1±3.1) with erectile dysfunction symptoms were assessed. Noninferiority safety was assessed in this retrospective, cross-sectional study with stratified random sampling by identification of prescribing in the presence of clinically important PDE-5 inhibitor drug interactions with or without high-risk cardiovascular disease, by asking about diagnostic symptoms specific to erectile dysfunction, and by determining frequency of patient counseling.

RESULTS

Noninferiority of the e-medicine system was shown for the 6 safety end points, relative to a traditional medicine system. Numbers of inappropriate prescriptions, after correction for disease and medication covariates, did not differ between systems. Medication counseling showed superiority of the e-medicine system. Standard diagnostic questions were required for e-medicine prescribing but were infrequently asked in traditional medicine.

CONCLUSION

Safety in prescribing PDE-5 inhibitors for erectile dysfunction was similar between a US-based, state-regulated Internet prescribing system and a multispecialty primary care system.

CI, confidence interval , IIEQ, International Index of Erectile Questions , PDE-5, phosphodiesterase type 5 , RR, relative risk

The Internet is rapidly becoming an important platform for health care communications.¹ This technological advance is driven by the delivery of health care from single to multiple physicians, by direct-to-consumer advertising that empowers patients to make their own health care decisions, and by greater public demand for rapid delivery of health care information.², ³

Unsurprisingly, the increase in demand for electronic health information has evolved in association with direct-to-consumer advertising of pharmaceuticals, which has led the public to seek Internet prescribing. However, prescribing via the Internet has resulted in legal, professional, confidentiality, and safety breaches that threaten public safety.⁴ In response to e-medicine prescribing, the health care industry has appropriately raised serious concerns about the safety of prescribing over the Internet.⁵, ⁶, ⁷

Two major obstacles exist for the safe and effective delivery of e-medicine prescribing. The first is need for a medical interview tool that generates a complete and accurate individualized medical history.⁸ An expert interview tool has been developed and is currently being used in an e-medicine prescribing system.⁹ The interview uses confirmatory questions and standardized scales to collect medical history, symptoms, drug history, and prior adverse drug events, providing a unique interview for each patient, to determine prescribing appropriateness.¹⁰ Information is standardized into a discrete, codified e-medicine record, reviewed by a licensed physician in good standing with a state regulatory agency and permitting direct customized patient consultation via electronic or telephone follow-up.

The second obstacle is consistent regulatory approval. One method to garner consistent regulatory approval would be for e-medicine businesses to grant access to their client databases in a transparent, verifiable, and confidential manner. This access would allow consistent accountability to federal and state regulatory authorities. The State of Utah entered into a consent order with KwikMed, an Internet prescribing company (Appendix¹¹), to permit e-prescribing of erectile dysfunction drugs under the Utah Practice Act, providing the welfare of the users remained paramount, that acceptable standards of care were consistently maintained, that properly supervised physicians and pharmacy personnel adhered to ethical conduct, and that monitors had constant access to Web site records while the users' confidentiality was maintained. Further, patients are provided direct access to the State of Utah for concerns through the Web site.

Phosphodiesterase type 5 (PDE-5) inhibitors were chosen for this study for several reasons. This drug class is safe and effective for erectile dysfunction, regardless of etiology, with clear contraindications. Further, recommendations for using the drug have been generally established by expert opinion, rather than by evidence, leading to variance in prescribing.

Use of an expert interview system for PDE-5 drugs with regulatory approval from the State of Utah suggests that e-prescribing via the KwikMed Web site is relatively safe. However, no independent data currently exist to assess safety of this type of prescription delivery system vs a traditional medicine system. Therefore, a stratified, random sampling technique was used to compare KwikMed e-medicine records to records from a traditional multidisciplinary primary care system from the Salt Lake City, UT, area to compare the noninferiority safety of prescribing PDE-5 inhibitors between the systems.

Back to Article Outline

PATIENTS AND METHODS 

A stratified random sample of 500 unique records from 102,734 e-medicine records and 110,000 traditional electronic charts from a multidisciplinary primary care system from January 1, 2001, to December 31, 2005, were selected. This database was searched for records of any patient with the diagnosis of erectile dysfunction. Data were assessed by degree of underlying disease burden for each patient in the database; 75% of the database was a consistent mix of patients, with 1 concomitant disease and receiving only 1 legend (requiring a prescription) drug. Fifteen percent of patients had a disease burden with 2 or more underlying diseases and were receiving 2 or more legend drugs. The remaining 10% had no underlying diseases and received no drugs. Similarly, 500 traditional medicine electronic records were then randomly selected. Because of the higher disease burden and subsequent number of legend drugs in the traditional medicine database, categories were set at 4 diseases for the largest portion of the database, more than 4 for the most disease burdened, and fewer than 4 for the least disease burdened. The disease burden was permitted to vary between databases to reflect the unique distribution of patients in each database.

Demographics, current and past medical history including alcohol use by patient history or by CAGE questionnaire (e-medicine only),¹² depression identified by patient history or by Centers for Epidemiologic Studies-Depression Scale (e-medicine only),¹³ and all medications including antihypertensive agents, nitrates, or α₁-adrenergic blockers were recorded. The history of erectile dysfunction was assessed by diagnostic questions about time of sexual difficulty, confidence in erection, erections acceptable for penetration, and erections maintained to complete intercourse.¹⁴ In traditional medicine, recording of these questions or a physical examination was noted if documented within 1 year of the incident visit for the PDE-5 inhibitor prescription. Each PDE-5 inhibitor prescription written or denied was analyzed, as well as any prescriptions contraindicated as determined by the package inserts for the 3 commercially available PDE-5 inhibitors. Prescription contraindications were determined by product label identifying serious drug interactions with PDE-5 inhibitors with or without presence of high-risk cardiovascular disease. High risk was defined as the presence of multiple risk factors with structural heart disease and a recent cardiovascular event. The contraindication was confirmed if the patient did not receive counseling. Patient education about PDE-5 inhibitor use, adverse effects, and avoidance of concomitant medications were analyzed. Renewal prescriptions and patient follow-up, including electronic, telephone, or face-to-face contact, were recorded for both systems. For traditional medicine patients, the mode of identification of erectile dysfunction was analyzed, including patient complaint, patient request, or telephone request.

This study was reviewed, declared exempt, and approved by the University of Utah Institutional Review Board.

Statistical Analyses 

Demographics, medical history, and drug history variables were compared between the e-medicine and traditional family medicine systems using standard statistical comparisons, or superiority tests, because differences were of interest. For categorical data with 2 categories, the groups were compared with the χ² test or Fisher exact test, as appropriate. For unordered categorical data with more than 2 categories, the groups were compared with the χ² test or the Fisher-Freeman-Halton test as appropriate. For ordered categorical data, the Mann-Whitney test was used. For continuous variables, groups were compared using a 2-sample t test with a Mann-Whitney test if data were skewed.

The safety comparisons of prescription appropriateness, frequency of validated International Index of Erectile Questions (IIEQs) for diagnosis of erectile dysfunction, and patient education were determined using a noninferiority test. Noninferiority studies test a hypothesis of whether a group is no worse than the comparison group.¹⁵, ¹⁶ In a noninferiority study, one group can be slightly worse than the other group, but without clinical consequence. This slight difference is referred to as the noninferiority bound, which is decided before the study is conducted. To determine whether e-medicine is just as safe as traditional medicine in prescribing PDE-5 inhibitors, a noninferiority study and statistical analysis were undertaken.

Noninferiority was inferred if the lower bound of a 2-sided 95% confidence interval (CI) for the risk ratio was greater than the relative noninferiority bound of 0.95. If noninferiority was demonstrated, then superiority was tested. This 2-stage testing of noninferiority and superiority controlled the type I error rate.¹⁷ To allow for multiple comparisons introduced by using 6 safety comparisons, the 95% CIs for the relative risks (RRs) were adjusted using Bonferroni multiple comparison procedure in the noninferiority comparisons.¹⁸ Thus, although “95% CI” is stated, these are actually Bonferroni-adjusted values for 6 comparisons, making them 99.17% CIs. This made the CIs wider and so more likely to cross or approach the noninferiority bound, correctly making it more difficult to demonstrate noninferiority. Confidence intervals adjusted using Bonferroni multiple comparison procedure were also used for the 6 superiority analyses. Superiority was inferred if the lower bound of the CI exceeded the null hypothesis value, RR = 1, which represents no difference. These wider CIs were more likely to cover or approach the null value, thus correctly making it more difficult to conclude a difference or superiority.

The selected sample size of 500 in each group allowed a 96% power to demonstrate noninferiority in both groups, assuming compliance on a safety comparison of 100% and 95%. Significance was set at P<.05 level for all tests, using 2-sided comparisons.

Because of the difference in age and disease burden between the 2 databases, a multivariate modified Poisson regression model was fitted to each variable listed in the Table to directly estimate the risk ratio, and then noninferiority comparisons were made using the adjusted risk ratios.¹⁹

TABLE 1. Demographics and Medical and Pharmacy History of e-Medicine Clients vs Traditional Medicine Patients^a

To balance the 2 delivery systems on underlying disease severity and medication frequency, the models were adjusted for age, body mass index, history of tobacco use, alcohol use, hypertension, coronary artery disease, heart failure, diabetes mellitus, liver disease, seizures, renal disease, pulmonary disease, priapism, Peyronie disease, abnormal curvature of the penis, prostate enlargement, depression, the number of legend drugs prescribed, nitrates, and α₁-adrenergic blockers.

For the diagnosis of erectile dysfunction using the IIEQs, the e-medicine system had a constant 100% compliance, with zero variance, as this was a required question in the system. To use a variance estimated from the traditional medicine system only, comparing compliance in traditional medicine to the constant of 100% provided a more generally acceptable analysis. This was accomplished by fitting a multivariate modified Poisson regression to traditional medicine only. Then the adjusted percentage of times the IIEQs were asked was estimated from the model, while holding the covariates constant at the total sample mean values, or combined group means. The risk ratio and CIs were then computed using RR = 1/(traditional medicine proportion), allowing a consistent presentation with the other outcomes of RR as the effect estimate. To clarify, because the e-medicine system had a fixed 100% compliance of asking the 4 IIEQs (required by the system design), its noninferiority was guaranteed without the need of a statistical demonstration, and proof of its superiority would have required only one patient who was nonadherent in the traditional medicine system. The statistical analyses were therefore used for quantifying the differences and providing CIs.

Back to Article Outline

RESULTS 

We selected 500 unique patients from e-medicine records and traditional family medicine electronic charts. Demographics for both groups are shown in the Table. Overall, the e-medicine clients were younger and healthier (demonstrated by a smaller body mass index and less disease burden) and were taking fewer medications.

The diagnosis of erectile dysfunction symptoms was assessed as one of the safety outcomes (Figure). The e-medicine clients were queried universally on all 4 IIEQs, whereas traditional medicine patients were queried less than 25% of the time. After adjusting for covariates in the Table, both noninferiority and superiority of the e-medicine system were shown for all 4 IIEQs: time of sexual difficulty (e-medicine, 100%; traditional medicine, 16.4%; adjusted RR, 4.9; 95% CI; 3.6-6.9), confidence in erection (e-medicine, 100%; traditional medicine, 17.2%; adjusted RR, 10.0; 95% CI, 7.3-13.8), whether erection was acceptable for penetration (e-medicine, 100%; traditional medicine, 9.8%; adjusted RR, 12.2; 95% CI, 8.1-18.4), and ability to maintain erection to complete intercourse (e-medicine, 100%; traditional medicine, 23.6%; adjusted RR, 3.6; 95% CI, 2.8-4.7).

• 
  • View Large Image
  • Download to PowerPoint

• FIGURE. 

  Noninferiority analysis of safety end points. CI = confidence interval; RR = relative risk.

The rate of contraindicated prescriptions was also a primary safety comparison. Eleven (2.2%) and 47 (9.4%) of prescriptions were contraindicated in the e-medicine and traditional medicine systems, respectively. Most contraindicated prescriptions occurred in the traditional medicine system when PDE-5 inhibitors were prescribed by telephone consultation. The system disparity was not evident, however, after adjusting for covariates in the Table. The 2 systems were shown to be comparable; when expressed in the same terms as other safety outcomes, 97.8% of e-medicine prescriptions were not contraindicated and 90.6% of traditional medicine prescriptions were not contraindicated (adjusted RR, 1.01; 95% CI, 0.96-1.06). Most of the difference was caused when physicians prescribed PDE-5 inhibitors by telephone call only (8.3% of the total prescribing in traditional medicine; 100% determined to be contraindicated). Superiority of one system over the other was not demonstrated on this safety outcome.

The third safety comparison was patient education (Figure). Written manufacturer product information was provided to all clients in the e-medicine system (100%). Tailored electronic messages were also sent to 75.2% of clients. In comparison, no medication instructions were recorded for 51.8% of family medicine prescriptions at the prescribing visit. After adjustment for the covariates, both noninferiority and superiority of the e-medicine system were demonstrated for this outcome (e-medicine, 100%; traditional medicine, 31.0%; adjusted RR, 3.4; 95% CI, 2.5-4.6).

A genital physical examination was frequently not recorded in the traditional medicine system. The recording rate of any physical examination was 67.6%. Genitals were examined for 19.8% of the patients. In 5 encounters, the patient refused a genital examination. The e-medicine records showed 80% of clients had undergone a general medical examination within the last year and 97% within 2 years.

It was more common in the traditional system to receive a PDE-5 inhibitor. A PDE-5 inhibitor prescription was filled at every incident encounter in the traditional medicine system vs 83% of the time in the e-medicine system by questionnaire only (an additional 12.4% were filled after telephone consultation between the physician and the client). In all e-medicine and traditional medicine encounters, patients had initiated the request for the drug or had experienced impotence, which led to the prescription. Sildenafil was the most frequently prescribed agent in both the e-medicine and traditional medicine systems, accounting for 73.6% and 90.4%, respectively (P<.001). Tadalafil and vardenafil were prescribed in 18.8% and 7.6% of the e-medicineclients vs 4.2% (P<.001) and 5.4% (P=.16) in the traditional medicine patients, respectively.

Follow-up in both groups was tracked by any query to the prescribed PDE-5 inhibitor, including tolerance, relief of symptoms, and prescription renewal or through change to another erectile dysfunction medication. Prescriptions were renewed at a 25% rate, independent of the system (P=.67).

Back to Article Outline

DISCUSSION 

Erectile dysfunction affects more than 100 million men worldwide, and more than 600,000 men aged 40 to 69 years seek care annually in the United States.²⁰, ²¹ Effective and reliable therapy with PDE-5 inhibitors is driving more men to seek treatment. Given the increasing use of the Internet to seek health care information and the social stigma of erectile dysfunction, the Internet is being increasingly used by men seeking erectile dysfunction treatment. However, the safety of these Internet prescription systems is appropriately being questioned because of lack of oversight by state regulation and the lack of perceived safety with the current face-to-face system.

This study demonstrated noninferiority safety end points of a state-regulated Internet e-prescribing system for erectile dysfunction medications compared with a traditional medicine system. Specifically, the e-medicine system demonstrated noninferiority to the traditional medicine system in the number of contraindicated prescriptions for PDE-5 inhibitors but was not superior. Prescription counseling occurred more often in the e-medicine system. The e-medicine system more frequently evaluated erectile dysfunction symptoms. The traditional medicine system treated a larger number of patients with erectile dysfunction who had a greater disease burden, and it more often targeted the physical examination to the primary disorder, rather than to erectile dysfunction.

Innovation, technology, and current medical practice all factor into the outcome of this study. Application of an expert interview system specifically targeted to erectile dysfunction along with a continuous platform for patient client-physician communications make this particular Internet system comparable to traditional medical practice. Medical practice guidelines for erectile dysfunction recommend a medical interview regarding the sexual dysfunction of the patient.¹⁴ Our results show that IIEQs are infrequently solicited from patients with erectile dysfunction, perhaps related to the social stigma of the disorder to both the patient and the physician or to the time constraints placed on physician visits. A complete multisystem physical examination, including a genital examination, should be conducted to reveal cardiovascular disease and neurologic and urologic abnormalities.¹⁴ The e-medicine system recorded a rate of 80% who had undergone a general medical examination within the past year. In comparison, physical examinations in the traditional medical system seemed to be conducted selectively, and a genital examination was often not performed. Diligence in traditional medical practices to conduct and to act on findings of the physical examination has been recently questioned.²²

Guidelines from the Federation of State Medical Boards for appropriate use of the Internet in medical practice and from the National Association Boards of Pharmacy Verified Internet Pharmacy Practice Sites state that a patient evaluation, including a physical examination to establish the diagnosis, must be documented before treatment.⁵, ²³ However, results of this study demonstrate that diagnostic questions and physical examinations directed to erectile dysfunction are often undocumented. These results were confirmed by a study of the 2002 National Ambulatory Medical Survey of office-based physicians not federally employed and by the National Hospital Ambulatory Medical Care Survey of hospital outpatient clinics and emergency departments analyzed for sildenafil prescribing.²⁴ In men 30 years or older who received a prescription for sildenafil, no diagnosis or complaint of erectile dysfunction was documented for half of them. The position of the Federation of State Medical Boards and National Association Boards of Pharmacy also appears to conflict with the American Medical Association's House of Delegates resolution adopted in 1999: “support[s] the use of the Internet to prescribe medications with appropriate safeguards to ensure that the standards for high quality medical care are fulfilled.” ⁶ Results of the current study suggest that, under a platform of regulatory oversight, the safety of e-medicine care can match or possibly exceed traditional medical care, providing an evidence-based framework to reexamine current guidelines.

In the e-medicine system, 80% of clients had had a general medical examination in the year before the encounter. Why did these patients not request treatment for erectile dysfunction in the traditional manner? Our results suggest that they would have received the drug if they had requested it. One may reflect that users of the Internet do not have to visit the physician's office, providing convenience. Online pharmacies can be more accessible to people with limited mobility and to people in remote areas.²⁵ Prescription authorizations can be securely placed 24 hours a day and delivered by mail.²⁶, ²⁷, ²⁸, ²⁹ This convenience is accentuated by potential safety; numerous errors occur when written or verbal instructions are misunderstood.³⁰ Providing standardized, consistent information and treatment that is transparent is possible for all patients with today's health care information technology, not selectively as in traditional medicine. Online pharmacies can reduce costs to patients for prescriptions because of leveraging economies of scale and decreasing transaction costs.³¹ Patients can avoid the social stigma, especially in small communities, of talking about erectile dysfunction with their physicians or of filling the PDE-5 inhibitor prescription at the local drugstore.

The e-medicine system was compared with a multidisciplinary primary care system for only erectile dysfunction medications. Because patients with erectile dysfunction are seen by a variety of practitioners, including from internal medicine, urology, and family medicine, the results of this study might not apply to all practices. Urologists document erectile dysfunction more often than family medicine, internal medicine, or other physicians.³² As it happens, the e-medicine system has a supervising urologist. Results of this study on e-medicine prescribing of PDE-5 inhibitors might not be universal for all medications.

The result of prescribing a PDE-5 inhibitor every time a patient with erectile dysfunction is encountered in the traditional medicine system requires discussion. Traditional clinicians might provide medication without considering contraindications. The electronic medical record in traditional medicine did not have a drug-interaction warning system, whereas it is available to the e-medicine system. Our finding that most contraindicated prescriptions were offered as a result of a telephone conversation with patients suggests that multitasking on the telephone without access to the electronic medical record could lead to errors. In the e-medicine system, prescribing decisions are always made with the electronic medical record as the focus. Another possibility is that clinicians who decide not to provide an erectile dysfunction medication do not document the diagnosis.

Two other limitations could be e-medicine clients' ability to truthfully answer interview questions and minimal time in the traditional system to complete documentation. Studies have found that patients are more honest in answering questions generated by a computer than in answering physicians' questions when face-to-face.⁸ In consideration of the current state of modern medical practice (eg, time pressures) and of the importance of comprehensive history recording to reimbursement, benefits of patients providing accurate patient histories through an interactive medical tool could apply across many practices, not be restricted to the Internet.

Back to Article Outline

CONCLUSION 

A state-regulated e-medicine system was shown to be similar to a traditional multidisciplinary primary care system for all safety end points in prescribing PDE-5 inhibitors. The e-medicine system outperformed the traditional system in most of the safety variables tested. Additional studies of e-medicine vs traditional medicine systems are needed to confirm our results.

We recommend that state regulatory agencies consider the regulatory model of oversight protections implemented by the state of Utah to license Internet prescribing companies.

Back to Article Outline

APPENDIX. E-Medicine Internet Web Site Procedures 

The Internet site prompts all clients to provide registration information, allowing the site to quickly identify new vs returning clients.¹¹ Registration information allows comprehensive oversight of potential client impersonation by cross-referencing registration information with the up-to-date database. Specifically, each new client name is matched to the database by name, with credit card information, occasionally through telephone contact, and, if necessary, by background searches on the Internet.

Each client is directed to an interactive patient interview questionnaire designed to collect a complete medical history for erectile dysfunction. The patented software assesses a client's mental health, determines the likelihood of alcohol problems, anxiety, and depression and collects history on all recognized contraindications and associated disorders for the permitted prescribing of legend drugs by the State of Utah Division of Occupational and Professional Licensing. The software consolidates the data into an electronic medical record reviewed by a physician registered and in good standing with the State of Utah.

Physicians can then make a decision as follows: (1) to prescribe for the patient an erectile dysfunction drug on the basis of the initial medical history provided through the patient interview questionnaire, (2) to collect additional information clarifying the medical history by either telephone or e-mail, or (3) to deny the prescription. If the prescription is approved, it is electronically sent for dispensing to a pharmacy registered in the State of Utah and to a pharmacist in good standing. Each medication is then shipped via an internationally recognized courier requiring an adult's signature for delivery.

If a client returns to the Internet site for refill authorization, a separate interactive questionnaire for refills (tailored to determine the patient's experience with the medication) must be completed. This refill questionnaire collects any new medical or drug information since the original prescription, before the same oversight of the prescribing and dispensing process transpires.

The electronic medical record created for each e-medicine client is transparent and contains a record of all medical data collected, all physician-client interactions, and all client administrative interactions. All clients are able to access their medical file online and see the resume of the prescribing physician and pharmacist. Patients are not allowed to alter their medical record but are able to alter their contact information. To provide continuity of care, patients may request a copy of their medical data be sent to their physician by providing the name and address to the system.

Back to Article Outline

REFERENCES 

1. Hesse BW , Nelson DE , Kreps GL , et al.   Trust and sources of health information: the impact of the Internet and its implications for health care providers: findings from the first Health Information National Trends Survey . JAMA . 2005;165(22):2618–2624
   • View In Article
2. HarrisInteractive. The Harris Poll Web site . www.harrisinteractive.com/harris_poll/index.asp Accessed June 24, 2008.
   • View In Article
3. Pew/Internet: Pew Internet and American Life Project Web site . http://www.pewinternet.org/reports/reports.asp Accessed June 24, 2008.
   • View In Article
4. Rost KT . Policing the “Wild West” world of Internet pharmacies . Chi-Kent L Rev . 2001;76:1333
   • View In Article
5. Federation of State Medical Boards of the United States of America  . Model guidelines for the appropriate use of the internet in medical practice . http://www.fsmb.org/pdf/2002_grpol_Use_of_Internet.pdf Accessed June 24, 2008.
   • View In Article
6. American Medical Association House of Delegates (A-99)  . Report of Conference Committee H . www.ama-assn.org/meetings/public/annual99/reports/hod/rtf/refcomh.rtf
   • View In Article
7. National Association of Boards of Pharmacy Web site  . http://www.nabp.net/vipps/intro.asp Accessed June 24, 2008.
   • View In Article
8. Bachman JW . The patient-computer interview: a neglected tool that can aid the clinician . Mayo Clin Proc . 2003;78(1):67–78
   • View In Article
   • Abstract
   • Full Text
   • Full-Text PDF (110 KB)
   • CrossRef
9. Wenner AR , Ferrante M , Belser D . Instant medical history . Proc Annu Symp Comput Appl Med Care . 1994;1036
   • View In Article
10. Dugaw JE , Civello K , Chuinard C , Jones GN . Will patients use a computer to give a medical history? . J Fam Pract . 2000;49(10):921–923
    • View In Article
    • MEDLINE
11. KwikMed Website  . http://www.kwikmed.com Accessed June 24, 2008.
    • View In Article
12. Ewing JA . Detecting alcoholism: the CAGE questionaire . JAMA . 1984;252(14):1905–1907
    • View In Article
    • MEDLINE
13. Radloff LS . The CES-D Scale: a self-report depression scale for research in the general population . Appl Psychol Measures . 1997;1:385–401
    • View In Article
14. Rosen RC , Riley A , Wagner G , Osterloh IH , Kirkpatrick J , Mishra A . The International Index of Erectile Function (IIEF): a multidimensional scale for assessment of erectile dysfunction . Urology . 1997;49(6):822–830
    • View In Article
    • Abstract
    • Full-Text PDF (695 KB)
    • CrossRef
15. Barker LE , Luman ET , McCauley MM , Chu SY . Assessing equivalence: an alternative to the use of difference tests for measuring disparities in vaccination coverage . Am J Epidemiol . 2002;156(11):1056–1061
    • View In Article
    • MEDLINE
    • CrossRef
16. Piaggio G , Elbourne DR , Altman DG , Pocock SJ , Evans SJ , CONSORT Group  . Reporting of noninferiority and equivalence randomized trials: an extension of the CONSORT Statement . JAMA . 2006;295(10):1152–1160
    • View In Article
    • CrossRef
17. Hung HM , Wang SJ . Multiple testing of noninferiority hypotheses in active controlled trials . J Biopharm Stat . 2004;14(2):327–335
    • View In Article
    • MEDLINE
    • CrossRef
18. Sankoh AJ , Huque MF , Dubey SD . Some comments on frequently used multiple endpoint adjustment methods in clinical trials . Stat Med . 1997;16(22):2529–2542
    • View In Article
    • MEDLINE
    • CrossRef
19. Zou G . A modified Poisson regression approach to prospective studies with binary data . Am J Epidemiol . 2004;159(7):702–706
    • View In Article
    • MEDLINE
    • CrossRef
20. NIH Consensus Development Panel on Impotence  . Impotence . JAMA . 1993;270(1):83–90
    • View In Article
    • MEDLINE
21. Johannes CB , Araujo AB , Feldman HA , Derby CA , Kleinman KP , McKinlay JB . Incidence of erectile dysfunction in men 40 to 69 years old: longitudinal results from the Massachusetts Male Aging Study . J Urol . 2000;163(2):460–463
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (55 KB)
    • CrossRef
22. Jauhar S . The demise of the physical exam . N Engl J Med . 2006;354(6):548–551
    • View In Article
    • CrossRef
23. National Association Boards of Pharmacy  . Verified Internet Pharmacy Practice Sites (VIPPS) Application Website . www.nabp.net/vipps/pharmacy/VIPPSApplication.pdf Accessed June 24, 2008.
    • View In Article
24. Young SE , Mainous AG , Diaz VA , Everett CJ . Practice patterns in sildenafil prescribing . Fam Med . 2006;38(2):110–115
    • View In Article
    • MEDLINE
25. Consumer advice: buying drugs online . Harv Health Lett . 2000;25(8):1–3
    • View In Article
    • MEDLINE
26. Henney JE . Cyberpharmacies and the role of the US Food and Drug Administration . J Med Internet Res . 2001;3(1):E3
    • View In Article
    • MEDLINE
27. Be WWWary of Internet drugs . Heart Advis . 2002;5(8):4–5
    • View In Article
    • MEDLINE
28. Hazlet TK , Bach MH . The Internet, confidentiality, and the . Camb Q Healthc Ethics . 2001;10(2):157–160 pharmacy.coms
    • View In Article
    • MEDLINE
    • CrossRef
29. Landis NT . Virtual pharmacies boast easy access, privacy safeguards . Am J Health Syst Pharm . 1999;56(12):1174; 1177-1179.
    • View In Article
    • MEDLINE
30. Martin MJ , Mayne JG , Taylor WF , Swenson MN . A health questionnaire based on paper-and-pencil medium individualized and produced by computer. II. Testing and evaluation . JAMA . 1969;208(11):2064–2068
    • View In Article
    • MEDLINE
31. Brushwood DB . Responsive regulation of Internet pharmacy practice . Ann Health Law . 2001;10:75–103
    • View In Article
    • MEDLINE
32. Beckman TJ , Abu-Lebdeh HS , Mynderse LA . Evaluation and medical management of erectile dysfunction . Mayo Clin Proc . 2006;81(3):385–390
    • View In Article
    • Abstract
    • Full Text
    • Full-Text PDF (53 KB)
    • CrossRef