As the AIDS epidemic continues to grow, women form an especially susceptible at-risk group; the majority of infections are through heterosexual transmission, and the risk of becoming infected during intercourse is greater for women than for men. In an effort to provide improved protection against HIV acquisition, women-controlled methods for defence against sexually transmitted infections and HIV, such as vaginally applied microbicides or devices, are increasingly the focus of interest and research. The first of these microbicides, nonoxynol-9, has proved to be disappointing in its ability to prevent infections.1 However, the search for an effective and acceptable microbicide continues, with a number of formulations at various stages of research development, including human testing.2
In addition to effectiveness, it is imperative to evaluate the safety of all new topical microbicides. As part of that evaluation, an examination of the colpos (cervix and vagina) to ascertain the presence of any physical or mechanical effects of such agents is recommended in a subset of human subjects participating in phase I and phase II studies of microbicides.3 Since the 1990s, such examinations have been performed by colposcopy. Although the clinical significance of all colposcopic findings remains unclear, breaks in the epithelium and obvious inflammatory or vascular changes are assumed to have relevance regarding a possible increased transmission of HIV.
Colposcopy is an examination of the cervix and vaginal tissues often through the use of an instrument known as a colposcope, which is basically a specially configured binocular microscope. However, in reality, a colposcope can be any instrument used to examine the tissues of the vagina and cervix through a magnifying lens.4 As such, an instrument that may not be traditionally thought of as a ‘colposcope’ can in fact be used both to inspect and record images of the cervix and vagina. In particular, a digital camera may be used to perform a magnified examination of the cervix and vagina. Digital assessment of the reproductive tract (DART) employs commonly available, ‘off-the-shelf’ instruments to perform colposcopy effectively. As such, it can be a practical and efficient means for studying the effects of microbicides on the vaginal and cervical epithelium.
In the context of researching microbicides in both developing and developed country settings, certain criteria must be considered in determining whether an instrument could be recommended for imaging the vagina and cervix in such research settings. First, the capacity for digital data storage is essential. In this way, images can be reviewed by outside observers and quality control can be monitored. In addition, the state of the epithelium of a participant can be followed more accurately over time. Because considerable research will be undertaken in areas with limited resources, a number of other factors relative to instrumentation must be considered. These include portability, ease of use, durability, and accessibility. DART fulfils these criteria and offers the potential for a cost-effective means of imaging the reproductive tract in the context of the development of safe microbicides and/or other women-controlled devices.
The core equipment required is a digital imaging device (camera) with at least a 3 megapixel capability. In addition, other features are essential for adequate imaging of the reproductive tract. The camera must have the capability for manual area or spot focus and manual spot metering. The ability to take timed/delayed shutter release images or have a remote ‘cable’ release device is also desirable in that it decreases the likelihood of motion artifact by touching the camera directly to record the image. The ability to connect to video display is necessary to magnify and visualize images in real-time adequately. The camera should have at least 3x optical zoom, a macro zoom setting and the ability to zoom digitally on a stored image in order to magnify images sufficiently. In addition, the capacity for larger (> 128 Mb) compact flash disks, memory sticks, or micro-drive storage is necessary as images taken require a significant storage space.
In addition to the camera, a light source to illuminate the vagina and cervix is essential. Although most digital cameras have an internal flash, in most cases the flash from the camera causes too much glare and shadows and does not allow real-time inspection of the tissue. Any light source which is ‘in front’ of the camera can be used. In our experience, the ‘KleenSpec’ disposable vaginal specula illumination system by Welch Allyn provides effective lighting (www.welchallyn.com). This halogen lighting system is plugged into either a 110 or 220 V electricity source and the bulb fits into a specially designed channel in the disposable specula. The lighting system costs US$227 in the United States. A disposable illuminator is also available from Welch Allyn for US$127 for a case (six illuminators per box, with six boxes per case). Each illuminator can be resterilized 50 times before disposal. The disposable vaginal specula are made of clear acrylic and are available in small, medium, and large sizes. They sell for approximately US$200 per case (100 specula per case). The specula are disposable, but can be processed and reused approximately five to 10 times. The clear, polystyrene specula that accommodate the expendable illuminator are available for US$105 a case (100 specula per case) and are available in small and medium Graves style. Other thin halogen lighting sources are also potentially useful.
The other essential piece of equipment is a ‘monopod’ stand to allow hands-free use of the camera, as well as the elimination of motion artifact. The monopod allows easier manipulation and takes up less space than a traditional tripod. Monopods with retractable ‘feet’ are available for approximately US$80.00.
The functionality of DART allows adequate imaging and has several other advantages, especially regarding its data storage and sharing capability. The storage capacity of a DART is excellent. Compact flash disks (or similar technology) are ubiquitous and becoming increasingly cheap. There is also excellent connectivity to computers. Because most cameras can directly connect to computers, electronic image transmission is easily accomplished. Alternatively, some cameras transfer images through a ‘reader’ to a computer, from which electronic transmission can occur.
The resolution of images with DART is adequate. In most cases images can be obtained that provide resolution comparable to 35 mm photographs, especially if the light is good and the camera can take a picture at 1 megapixel or greater. If the setting is set at least to ‘fine’ or 1 megapixel, then ‘zooming in’ to a level that is equivalent to approximately 6x on a colposcope can be accomplished that provides decent resolution. When using a camera as the visualizing and recording device, the technology cannot produce the type of binocular vision obtained with traditional colposcopes. However, even with the best colposcopes, the advantages of three-dimensional binocular vision are reduced to two dimensions when a picture is taken and the image subsequently recorded or printed.
The image seen with the digital camera in the DART system can be displayed on a larger video screen to view the images in real-time. However, in the specific case of digital still cameras, images are of better quality in ‘playback’ mode, which can also be viewed at the time of examination on the video screen. This also allows the immediate gratification of checking image quality in real-time. If an image is not satisfactory, it can be retaken at the time of examination rather than wasting film or potentially missing the opportunity for adequate recording. When using digital video cameras, depending on the quality of the camera, the quality of the actual real-time video image can be excellent, but the capacity to capture and store a high resolution still image is variable. At present, digital cameras may represent a better value-for-money option than video cameras.
When used in the field, DART must meet several important programmatic criteria. Because microbicide research is performed in both developed countries and developing countries, DART must be feasible for a wide range of settings and resource levels.
As mentioned earlier, a good light source is critical. However, because several good options are available, this is a manageable issue. The possibility of a battery-powered light source would add mobility and utility to the DART system.
DART may be an especially useful method of visualization in the field because of the durability of its components. Digital cameras are generally very stable and require almost no calibration or adjustments. In addition, digital cameras are highly portable. Even the highest resolution cameras can fit in a small camera case. Digital cameras are also quite accessible, as they can be purchased almost anywhere. Most manufacturers have regional distributors, where maintenance and repairs can be provided.
The relative inexpensiveness of digital cameras (with the technology becoming increasingly inexpensive) makes it feasible to have a spare camera around in case of breakage. In case of loss, it is much easier to obtain and ship to the field a new camera compared with a new colposcope (Table 1).
The power source for digital cameras is usually provided by batteries, either AA or proprietary types. Rechargeable batteries are recommended. Cameras can be recharged either directly with an AC adapter, or with a separate charging device. The batteries can be recharged almost indefinitely. Most batteries will last a clinic session or two, depending on how many photos are taken and at what resolution. Higher resolution pictures consume more power.
The ‘mailability’ of the images produced makes DART especially useful. Images can be ‘mailed’ electronically either by e-mail or, if images are large, by the use of a file transfer (ftp) site. Images can also be stored on any of a variety of media and then mailed on the media such as CD-Rom or USB drives. Also, case record forms (CRFs) could be electronic and the digital images could be ‘pasted’ directly on to the CRF. The CRF could then be printed out and filled in by hand, or filled in on the computer directly. A central source of electronic storage could then be used. In addition, quality assurance is enhanced as the images can be reviewed directly.
Although no studies have directly compared the results of a digital assessment of the reproductive tract with a traditional colposcopic assessment, there is considerable evidence that digital camera-based assessments are reasonably consistent and accurate. For example, several colposcope manufacturers now provide accessories by which a digital camera can be attached to a colposcope for recording the image seen through the colposcope. However, when comparing images recorded directly from the (hand-held) digital camera with those obtained by using a colposcope with an attached digital camera, it is difficult to tell the images apart (Figure 1). In addition, recognizing that a ‘video’ colposcope (basically a video camera with an internal light source) is as accurate as its optical counterparts (i.e. the traditional colposcope) in identifying dysplasia, there is reason to believe that the still version (i.e. a digital camera) would be dichotomously (yes or no, present or not present, she does or she does not) just as capable of identifying the kinds of lesions described in the CONRAD/WHO guide to the assessment of the reproductive tract. In sum, DART probably represents a practical, cost-effective means of assessing the effects of the microbicides and/or other devices on the genital epithelium of both women and men.