Efficacy is the improvement that can be achieved in a desired health outcome with use of an intervention in a research setting by expert subjects under ideal conditions.

Effectiveness is the improvement that can be achieved in a desired outcome with use of an intervention in a target community (or by an individual in the "real world") under resource constraints and with typical use.

Method failure is the failure of a protective device that is dependent on the physical properties of the device. Note: with regard to the latex condom, combined slippage and breakage rates are commonly referred to as method failures. Another form of method failure occurs when a condom is used consistently and correctly but infection is transmitted by direct contact with infected skin or skin lesions that lie outside the areas covered by a male condom (vulva, scrotum, thighs, etc.)

User failure is the failure of the device due to incorrect and/or inconsistent use. Note: slippage during intercourse is usually considered a method failure, but slippage during withdrawal is considered a user failure.

Perfect use is obtained by using a condom for every act of intercourse and following the recommended steps for correct condom use during each use. Note: only one of the studies reviewed mentioned both "correct" condom use and condom use during every act of intercourse.

Always use is using a condom during every act of intercourse.

Never use is not using a condom during any act of intercourse.

Typical use is using a condom more often than never use but less often than always use. Typical use includes consistent and inconsistent use, as well as correct and incorrect use.

-- Modified on 1/30/2003 2:17:09 AM

RISK REDUCTION VERSUS PREVENTION

The language used to describe the potential benefits of condom use can be confusing. Feldblum and Cates make the following statement: "Product labeling and counseling of people at risk have to make a clear distinction between absolute protection ('prevents infection') and partial protection ('reduces the risk of infection').

Given the ever-present risk of infection due to method failure alone, one should not expect condom use to prevent infection. Risk reduction may be a legitimate expectation, but the amount of risk reduction is influenced not only by user failure (inconsistent and incorrect use) and the mode of disease transmission, but also by other disease-specific factors such as the "disease-specific infectivity," or DSI.

CUMULATIVE RISK

Cumulative risk is the likelihood of an outcome occurring at least once, given a repeated number of risk exposures. Feldblum and Cates say this about cumulative risk: "Another important counseling point is that single-episode (condom) efficacy and the cumulative (condom) efficacy diverge widely as the number of exposures to an infected person increases." For example, an intervention that is 99.8% effective for a single episode of intercourse can yield an 18% cumulative failure rate with 100 exposures.

The following list contains additional factors that influence whether a condom will effectively reduce the risk of transmission (or acquisition) of an STD.

MECHANICAL QUALITIES OF CONDOM MATERIALS

An intact latex barrier should contain the infectious agents for all known STDs that are transmitted by the exchange of infected bodily fluids. Some studies show leakage of virus-sized particles through intact latex barriers, but actual infection from such leakage is unlikely, and this leakage is probably not a significant factor in disease transmission.

MODE OF STD TRANSMISSION

For STDs spread by genital fluids (HIV, gonorrhea, chlamydia, and trichomoniasis) an intact latex condom should prevent one's risk of infection following a single sexual exposure -- provided the condom is used according to the seven steps for correct use recommended by the CDC and does not slip or break. For STDs classified as genital ulcer diseases (genital herpes, syphilis, and chancroid) and for HPV (which appears to be transmitted both by direct contact and through genital fluids), an intact latex condom may not reduce one's risk for a single act of sex, even when the condom is used "correctly" and does not break or slip. Despite even correct condom use, significant risk of infection may remain because disease transmission can occur from infected skin or lesions not covered by the condom.

METHOD FAILURE (SLIPPAGE AND BREAKAGE)

Even when a condom is used consistently and correctly, slippage and breakage can occur. Sudies demonstrate that slippage or breakage occurs between 1.0% and 3.6% of the time. The reported slippage and breakage rates are even higher for less experienced users.

A large French national survey reported differences in slippage and breakage rates between couples with less than 5 years of condom experience and couples with more than 5 years of experience. Couples with greater than five years of condom experience reported 0.8% breakage and 0.7% slippage, for a total method failure of 1.5%. Those with less than five years of condom experience reported a 6.2% breakage rate and a 1.8% slippage rate, for a total method failure of 8.0%.

Condom breakage and slippage can also result in pregnancy. In a large study of over 17,000 women, a difference in condom failure rates was determined to depend on the number of years of condom experience. Among married women ages 25 to 34 years, the pregnancy rate for those with less than 2 years of condom experience was 6.0%. When the length of condom experience increased to 25 to 48 months, the pregnancy rate declined to 4.0%, and with 49 months of condom use, the pregnancy rate was 3.6%.

USER FAILURE

The form of user failure that carries the greatest risk of infection is clearly condom nonuse. If condoms reduce the risk of contracting/transmitting any STDs, that benefit is lost or substantially diminished when condoms are not used for each and every act of intercourse. In addition to inconsistent use, condom effectiveness is also compromised by incorrect use. Incorrect use means not performing all the recommended 7 steps for correct use before each act of intercourse.

Incorrect use can take many forms, for example, initiating a genital contact before putting on a condom or using a condom with a petroleum-based lubricant. Another potential form of incorrect use is the practice of "reversing" or flipping the condom over after having first placed it on the penis with the wrong side down. Improper initial positioning of the condom makes unrolling the condom difficult. When initial attempts to unroll the condom fail and the condom is flipped over, any secretions -- including sperm-containing pre-ejaculate and/or infectious fluids -- that remain on the condom may be in direct contact with the vagina and cervix during vaginal intercourse.

STD INFECTIVITY

Infectivity means "infectiousness" and is a measure of one's risk of infection given an exposure to an infectious disease. Infectivity is sometimes expressed as a "per-parnership infectivity" and at other times as "per-act infectivity." Different sexually transmitted infections have different infectivities. The reported STD infectivities range from 0.001 for HIV to 0.70 for chancroid. (Note: the infectivity of HIV varies with the stage of the disease, treatment status, and other variables.)

Infectivity varies not only by disease but also, for a given disease, often by gender. For example, with gonorrhea, males have a 0.20 infection risk per act of sexual intercourse with an infected partner, but females have at least a 0.50 risk of infection per act with an infected partner. Condom failure is more likely to result in infection when that failure exposes one to a particularly infective STD.

PRESENCE OF OTHER STDS

The presence of another STD -- paricularly, an ulcerative disease -- can significantly increase the risk of acquiring (or transmitting) HIV. It is unknown whether the presence of a non-HIV STD can increas the risk of acquiring (or transmitting) another non-HIV infection.

AGE AND SEX

The colummar cells covering large exposed portions of the young adolescent cervix are more susceptible to certain infectious agents -- including HPV, Neisseria gonorroeae, and Chalmydia trachomatis -- than are the squamous epithlial cells that typically cover most of the adult cervix. When adolescent females with this cervical condition ("cervical ectopy") come in contact with these infectious agents, infection is more likely.

NUMBER OF EXPOSURES

Peterman points out that the total number of sex acts (exposures) is an important variable in determining condom effectiveness. This concept is supported by the mathematical modeling work of Mann et al and is discussed further in the following section on Cumulative Risk.

-- Modified on 1/30/2003 2:24:36 AM

The issue of cumulative risk was not addressed in the NIH panel's report. Cumulative risk is the likelihood of an outcome occurring at least once, given a repeated number of risk exposures. How effectively a condom can reduce one's infection risk for a single act of sex with an infected partner is important, but what happens to the risk of infection after repeated exposures -- even for "always" condom users -- helps determine whether the uninfected partner ultimately becomes infected.

Mann et al, for example, show that if one assumes a 3% slippage and breakage rate and a disease-specific infectivity of 0.50, then the expected risk reduction for one act of sex with an infected partner while using a condom correctly is 98.5%. The calculated cumulative risk of condom failure is 14% for 10 acts of sex with an infected individual, 26% for 20 acts of sex, and 37% for 30 acts.

If incorrect condom use is added to the likelihood of condom breakage and slippage, the risk of exposure and subsequent infection is also higher. According to Warner et al, in a population of self-selected well-educated men who average more than 5 years of condom experience, 10% of all condom uses left the male user with a potential risk of infection (slippage, breakage, and incorrect use). The authors concluded: "Given the level of exposure to unprotected intercourse during condom use in this population, we suspect that exposure levels may be even grater in less-experienced, less-educated populations, such as new condom users and adolescents."

...

-- Modified on 1/30/2003 2:27:59 AM

The NIH Condom Effectiveness Panel examined the degree of risk reduction afforded by condom use for each of eight common STDs. The results from the three cohort studies that measured aggregate outcomes for multiple infections were generally excluded. An exception was the inclusion of selected data about chlamydia from the Zenilman study. We feel that the findings of these studies should be discussed here.

The cohort study reported by Zenilman et al was a prospective investigation conducted in Baltimore STD clinics. Extensive sexual histories were obtained from all study subjects, who additionally were evaluated for the presence of an STD. When identified at the intake examination, existing nonviral STDs were treated. Subjects were then followed for the development of four incident STDs (gonorrhea, chlamydia, syphilis, or trichomoniasis). Those who were infected at the 3-month visit were classified as having a new infection.

Of those with new infections, 15% of the men who "always" used condoms were newly infected, compared with 15.3% of "never users." twenty-three and one half percent of women who reported that their male partners "always" used condoms had a new STD, compared with 26.8% of women who said their partners "never" used condoms. These differences in the development of new infections between the "always" and "never" condom users were not statistically significant.

The authors gave three explanations why their study failed to demonstrate a significant relationship between self-reported condom use and new STDs, but none of these explanations mentioned the possibility of condom failure (breakage or slippage) or the cumulative risk of infection as an explanation. Nor were these issues mentioned in the many letters to the editor that followed publication of the article.


-- Modified on 1/30/2003 2:33:42 AM

Many of the published, peer-reviewed articles reviewed by the NIH Condom Effectiveness Workshop Panel that form the basis for this information have been available for some time. Nonetheless, many of the issues and concerns identified by the panel are "new" to both the public and to clinicians who counsel patients in their practices. Efforts to communicate this "new" and vital information -- the proven benefits, known risks, and unanswered questions -- to the public and to health professionals must now begin.

On the basis of this information, the authors suggest the following:

(1) In the absence of compelling evidence of condom effectiveness, young people should be strongly encouraged and counseled to delay the initiation of sexual activity. Sudies have shown that delaying the onset of sexual activity markedly decreases the average number of life-time sexual partners. The number of lifetime sexual partners greatly influences the risk of STD acquisition. Delaying sexual initiation is paricularly important because of the high incidence and prevalence of STDs in the adolescent population. Though this delay in sexual initiation will benefit both sexes, females are likely to receive the most significant health benefits. Allowing time for the epithelial cells of the female cervix to mature before any potential STD exposure occurs should provide measureable protection.

(2) Patients using condoms should be provided with information to explain the difference between absolute protection (which prevents infection) and partial protection (which reduces the risk of infection). Condom users must know that condoms, at best, reduce the risk of some STDs and may or may not reduce the risk of other STDs.

(3) Patients using condoms should be provided with the information that their risk of contracting an STD increases with the number of disease exposures -- even when condoms are used consistently and correctly. This increase is even more dramatic with inconsistent or no condom use.

(4) Patients using condoms should be provided with the information that condoms are not equally protective for every STD. The actual degree of risk reduction provided by the condom depends on other factors, many of which are outside the user's control, including gender, mode of disease transmission, and disease-specific infectivity. Condoms appear to be "more forgiving" for STDs with low infectivities (such as HIV) than for STDs with high infectivities (such as gonorrhea in females).

(5) Patients using condoms should be provided with information about avoiding high-risk sex partners. Having sex with a prostitute or IV drug user is clearly a high-risk activity. Similarly, having sex with an individual who has (or has had) multiple sex partners is also a high-risk activity.

(6) All patients should be provided with the information that STDs often are asymptomatic and go unrecognized, even by those infected. Patients should understand the necessity of appropriate screening, even for consistent condom users.

(7) All patients identified as having an STD should be counseled that reinfection is common after STD treatment. To lower reinfection rates, one's partner must also be tested and treated. In addition, posttreatment testing to confirm cure may be indicated for certain STDs.

(8) Healthcare providers should actively identify patients in their practices who are engaged in risky behaviors, who have partners who admit to engaging in risk behaviors, or who, because of other characteristics, are members of a risk group. Once these individuals are identified, healthcare providers should systematically test them for STDs. Even patients who report "always" condom use should receive screening and/or testing. This requires healthcare providers to implement systems for collecting information about the sexual activity and risk status of each patient at every office visit and for counseling against involvement in risky sexual behaviors.

-- Modified on 1/30/2003 2:39:35 AM