Visualization of manufacturing defects in latex gloves and condoms.

Visualization of manufacturing defects in latex gloves and condoms

Concern over the transmission of the human immunodeficiency virus (HIV), either through sexual activity or exposure of health care professionals to blood and other body fluids of patients, has led to a heightened interest in the effectiveness of latex barrier products such as gloves and condoms. Both the manufacturers and regulators of these products have been reexamining quality assurance procedures and tests in light of this concern (ref. 1).

Protective barriers reduce the risk of exposure to potentially infective substances, but they can fail for several reasons:

* flaws in the manufacturing process which result in defective products;

* breakage of an apparently non-defective product during use;

* failure to use the barrier properly; and

* failure to use the barrier during every potential exposure. This study examines only the first of these classes of failures - manufacturing defects.

Many manufacturers of latex surgical and examination gloves check a fraction of production (up to 100% in some cases) in an air-leak test before packaging, and a sample (up to 100% in some cases) of finished product is periodically checked with a water-leak test before lot acceptance. Nearly all manufacturers subject condoms to 100% electrical testing for all production, and samples of the finished product are usually air-burst and water-leak tested. Defects are almost inevitably produced during manufacturing, and these tests are needed to assure quality in the finished products.

For both condoms and gloves, many manufacturers conduct destructive materials testing procedures on samples of latex from finished products. However, these tests are not normally intended for the detection of the types of defects discussed in this article.

The Food and Drug Administration (FDA) regulates the manufacture and introduction into commercial distribution of latex barrier products. It does so by inspecting manufacturing plants and checking for compliance with latex product regulations in randomly sampled lots. FDA recently issued a proposed rule that requires latex surgical gloves to have a defect rate of no more than 2.5% in the finished product (ref. 2); the defect limit for patient examination gloves is 4%. Condoms must meet a somewhat more stringent standard of 0.4% (ref. 3).

To characterize the nature of the defects which lead to a glove or condom being rejected by a test for leakage, the present study examined a group of products which had, in the case of condoms, failed the electrical screening test or, in the case of gloves, had failed a water-leak test. Photomicrographs of the defects were made and examined in an attempt to determine the likely cause of the defect. Such information may contribute to improvements in the manufacturing processes.

It should be noted that only two manufacturers, one of gloves and one of condoms, are represented in the products we examined here, so the defects characterized may not be representative. However, the latex dipping processes employed by these companies are fairly representative of the industry, and these results are probably similar to those which would be obtained for other manufacturers.

Materials and methods

A Zeiss IM 35 microscope, outfitted with a Contax Model 137 35 mm camera, was used for all photomicrographs. Various microscope objectives (x3.2, numerical aperature 0.07; x10, NA 0.22; x40, NA 0.75) were selected, depending on defect size. Either Kodak TMX 5052 or TMY 5053 black/white film was used.

The thicknesses of the gloves and condoms were measured with a dial micrometer with spring-loaded 10 mm contacts (Mitutoyo, model 2804-10). Measurements of the thickness of gloves in the palm area ranged from about 165-175 [mu] and about 200-230 [mu] in the fingers. Condom thickness ranged from about 100 [mu] near the tip to about 50 [mu] near the rim.

Gloves

Two separate sets of latex surgical gloves that had failed a manufacturer's water leakage test were obtained:

* a set of 12 with unspecified defects;

* a set of nine, consisting of three groups of three each, according to the site of defect (finger tip, finger crotch and palm).

Set 1: Since the 12 gloves had failed the manufacturer's water leak test, and the exact defect area had not been identified, each glove was first filled with water to determine the exact site of the defect and to verify whether or not the glove actually leaked. The outer and inner surfaces of the identified defect areas were then photographed and measurements were taken.

Set 2: Photographs (of inner and outer glove surfaces) and measurements were taken of the manufacturer-identified defect areas in the nine gloves.

Note that "inside," at least in the case of the gloves, refers to that side of the glove normally in direct contact with the hand.

Condoms

Forty-four condoms (non-lubricated, receptacle end) that had failed a manufacturer's wet electrical test were obtained. This was not a random sample of all condoms failing their electrical test. The manufacturer selected only those for which some defect (not necessarily a hole) was visually observed; however, neither the nature nor the cause of the defects were identified by the manufacturer.

In our laboratory the 44 condoms were examined visually and microscopically. A complete visual examination was done only for the outer surface; the outer and inner surfaces of the defect areas in more than half of this sample (30 of the 44) were also photographed. All 44 condoms were then subjected to a 300 ml water-leak test to verify the suspect defect area as well as to ascertain whether the holes detected by the electronic test did, in fact, leak.

A condom support sleeve was fabricated and held by a ring stand. Condoms were mounted onto the sleeve so they would hang freely. The condoms were then filled with 300 ml water and checked for leaks. Where no leak was evident on hanging, the condom was removed and the open end closed. It was laid on its side and rolled gently back and forth on an absorbent towel (brown paper, in this case). If leaks were still not evident, pressure, which tended to stretch any holes, was applied during rolling. The water-leak/rolling test also had the ability to detect leakage outside of the identified defect area.

Results

Gloves

The nature and size of defects in both sets of gloves are illustrated in the following figures. A summary of glove defects is found in table 1.

Table : Table 1 - glove defect summary

Glove   Sample   Bubble   Inclusions   Film defects
 set     size    burst     closed      or fractures
  1       10       11        0               3
  2       9        8         3               2

(Some samples had more than one defect)

Set 1: Leaks were detected in 11 of the 12 gloves, including one which only had a large tear near the cuff (which would have automatically led to its rejection; no other leak in this glove could be found). No leak could be found in one of the gloves, even though the manufacturer had supposedly identified one. It was filled with water on three separate occasions, squeezed, and subjected to an air pressure test, but with no evidence of leakage. This left 10 gloves in this set to examine. Since several gloves leaked from more than one area, there were a total of 14 defects to examine photographically. Holes in gloves ranged approximately in size from 20 to 1200 [mu]. In addition, a defect in one glove consisted of an extra fragment of latex, measuring 2.4 x 0.9 mm, attached to the surface. In most gloves in this set, the hole diameter was different on the inside of the glove surface from that on the outside surface, with the inner surface dimension generally being larger.

The most common defect appeared to be an air bubble in the latex film that perhaps had blown out during leak testing. The surface of the bubble may have opened up during the curing process, leaving only a thin intact film over the surface next to the form, and a spherical void within the film, as shown schematically in figure 1, and photographically in figures 2 and 3. The inside of the glove can be distinguished from the outside from the roll of the cuff. We also saw air bubble clusters, as illustrated in figure 4. The gloves would rupture through the remaining thin membrane either when tested, or possibly upon removal from the form.

Set 2: This glove set was pre-selected by the manufacturer to include a distribution of defects in the finger tip, finger crotch and palm areas. Bubble inclusions were found to be the major cause of holes, and were observed in all three areas of the latex membrane in these samples. These inclusions were the main defect found in the palm area. Some inclusions resulted in the typical "burst bubble" type of hole as shown in figure 5. However, most of the holes in this set had inner and outer surface dimensions that were approximately equal and circular on both surfaces. A non-leaking bubble defect where the thin wall of the bubble had not burst was observed in three gloves in one finger tip and two palms. Defects in the finger tip area were the most varied. These included one of a "burst-bubble" type (shown in figure 5); another was a defect that led to an apparent fracturing of the membrane; and another defect consisted of an irregular opacity.

Choosing the smallest and the largest dimensions, hole sizes in this set ranged from 20 to 1200 [mu] for bubble defects, to 20 to 1000 [mu] for film fractures. The majority of holes/defects in this set of gloves were large (i.e., in the 200 to 1200 [mu] range).

Condoms

Of the 44 condoms examined visually and microscopically, defect areas in 30 were photographed and measured. All condoms were then leak tested. Most of the defects were visibly detectable with the unaided eye, though in some cases the defect was just a small speck in the latex, which might have been difficult to see without careful examination. It appears that approximately one-third of the "defects" were caused by a scratch or snag ("abrasion") of the condom surface, which, in many cases, left fairly large cuts/tears. Thirty-eight of the 44 condoms contained visible white or black specks ("inclusions") in the latex. All inclusions appear black in the photographs since the photographs were taken by transillumination. A summary is found in table 2. Most of the condoms readily failed the leak testing procedure; however, it is possible that several would have passed, depending on the technique of the tester. We had to apply pressure to force water out of several of the condoms. The only detected leaks were from the specifically marked defect area. However, in cases where there were major tears, leak testing of the remainder of the condom was sometimes problematic. Although particle inclusions were sometimes found outside of the marked defect area, none of these leaked.

Table : Table 2 - condom defect summary

Sample     Particle                 Other
 size     inclusions   Abrasions   defects
  44          38          14          3

(Some samples had more than one defect)

Some representative examples of defects found include: snag (figure 6); cut/tear (figure 7); inclusions (figures 8 and 9); and heterogeneous latex flaw (figure 10).

Discussion

It should be emphasized that we only examined latex products which had been found by the manufacturer to be defective. Our results do not address the question of the rate of defects in the finished product.

Defects in latex products manufactured by dipping processes are almost impossible to avoid completely, and these defects must be discovered by the manufacturer's quality assurance procedures. Defects we have observed in these samples include air bubbles, particulate inclusions, abrasions, thin spots and extra bits of cured latex in the film. In gloves, over half of the defects appeared to be primarily due to air bubbles which presumably formed during the dipping/curing process. The condom defects appear to be divided between dipping/curing problems such as particulate inclusions, and handling problems such as tears, snags and abrasions. About three-fourths of the latex glove manufacturers use a single-dip process. This method is intended to produce a viable barrier in a single film of latex.

The gloves examined in this report were from a single-dip process and it is not surprising that the majority of defects are characteristic of the limitations of this process. The rationale for a double-dip process is that air bubbles or other defects would rarely coincide in the two layers, and would be less likely to result in leaks. That does not mean that products from a double-dip process are necessarily better than others, but it does mean that the testing of the single-dip products may need to be more thorough, and, though the evidence is lacking, may result in more rejected production than a double-dip process.

Most condoms are manufactured with a double-dip process, and as expected, the prevalence of the different types of defects are somewhat different than for the gloves. The air bubble defects were uncommon in condoms, while snags and tears - defects characteristic of post-dipping handling problems - were much more common in condoms than in gloves. The condoms did have a number of defects due to particulate inclusions which led to their rejection by the electrical leakage test (not all of which actually leaked).

Manufacturing latex products with the dipping/curing approach will probably never be a perfect technology and quality assurance procedures will always be necessary to weed out defects. However, controlling the processing parameters carefully could minimize bubble and inclusion defects, and careful examination of handling procedures after dipping could minimize abrasions.

References

1. Carey, R., Herman, W., Herman, B., Krop, B., Casamento, J. (1989), A laboratory evaluation of standard leakage tests for surgical and examination gloves. Journal of Clinical Engineering, Vol. 14, No. 2:133-143. 2. FDA proposed rule - Medical devices; patient examination and surgeons' gloves; adulteration. 21 CFR Part 800, Vol. 54, No. 223, November 21, 1989. 3. Compliance policy guide #7123.21, Condoms: Defects - criteria for direct reference seizure, Dec. 30, 1987.

PHOTO : Figure 1 - schematic of burst bubble latex defect

PHOTO : Figure 2 - example of burst air bubble

PHOTO : Figure 3 - example of burst air bubble

PHOTO : Figure 4 - example of air bubble cluster

PHOTO : Figure 5 - example of burst bubble

PHOTO : Figure 6 - example of snag defect in condom

PHOTO : Figure 7 - example of cut/tear defect in condom

PHOTO : Figure 8 - example of condom inclusion defect

PHOTO : Figure 9 - example of condom inclusion defect

PHOTO : Figure 10 - example of heterogenous latex flaw defect in condom