A blend of Midsouth cotton of predominantly grade code 52 (i.e., color grade code 52 and leaf grade code 5 according to the current classification system [USDA 1993a]) was selected as a worst-case test. This is the lowest grade that, following a continuous process wash, would be exempted under the 1985 revised standard from all provisions except for medical surveillance and medical recordkeeping. Furthermore, over a recent 5-year period, less than 10% of the cotton crop was lower than this grade [USDA 1989, 1990, 1991, 1992, 1993b, 1994]. Part of the blended cotton was left unwashed, part was mildly washed using a continuous-wash system according to process criteria stipulated in the 1985 revised OSHA standard, and part was mildly washed in the modern batch kier systemtwo kiers each of two different washing conditions. One of the batch kier wash conditions involved temperature and water-to-fiber ratio as specified in the 1985 revised standard for continuous wash systems. The other wash conditionlower water-to-fiber ratio and shorter wash time, but with a higher temperaturewas considered a priori to be more commercially feasible.
Compared with the unwashed cotton, all three mild washings resulted in substantial and statistically significant reduction of card-generated airborne cotton dustin each case by at least 50%. In addition, the three different mild wash treatments were highly effective and statistically equivalent in reducing the endotoxin content of card-generated airborne elutriated dust. As a result, the concentration of airborne endotoxina more specific indicator of the respiratory hazard contained in cotton dust than gravimetrically measured dust concentration (see Appendix)was very effectively reduced by all three washings, from more than 300 ng/m3 for the unwashed cotton (at a dust level of 1.98 mg/m3) to less than 10 ng/m3 for each of the washed cottons (at dust levels ranging from 0.35 mg/m3 to 0.89 mg/m3). These low airborne endotoxin levels generated during card processing of the washed cottons were all below a relative "threshold" for acute airway response in humans described previously by NIOSH investigators in this same setting [Castellan et al. 1987]. Given the well-recognized interlaboratory variability associated with measurement of endotoxin activity in dust samples [Perkins 1992; Chun and Perkins 1994], it is important to note that the endotoxin assays were from the same NIOSH laboratory using the same standards that were used in the earlier studies.
In sum, the results of this research clearly demonstrated that the modern batch kier system is capable of mildly washing cotton as effectively as the already partially exempted continuous batt process. However, after reviewing the research results and obtaining input from selected Task Force member organizations, OSHA was disinclined to provide any new exemption for batch kier washed cotton without first having "additional data from washing trials involving a range of cotton blends at more than one washing facility," as well as more information on implications for chronic respiratory disease potential of mildly washed cotton [Adkins 1990]. Over the past several years, research findings have been generated that provide this additional information.
On the basis of human ventilatory responses to experimental exposures to dust from this washed cotton, Jacobs and colleagues concluded that their results "suggest that modern batch-kier systems can effectively remove the acute pulmonary toxicity of cottons washed at 60°C and a 40:1 water-to-fiber ratio" [Jacobs et al. 1993]. The endotoxin assay results from the NIOSH laboratory on card-generated airborne dust samples support this overall conclusion. Current evidence offers a basis for accepting the measurements of acute human airway response and airborne endotoxin made during the washing studiesnot only as indicators of potential acute respiratory hazard, but also as surrogates for the chronic respiratory hazard of cotton dust (see Appendix).