A Comparative Analysis of the Risk of Decompression Sickness with respect to Dive Profile and Associated Dive Depth, Ascent Rate, Dive Era, and Inspired Gas Composition

Abstract

To date, no systematic review has been conducted to assess the compounded risk of decompression sickness (DCS) with respect to more nuanced variables such as dive profile (dive depth, ascent rate), dive era, and inspired gas composition. The aim of this study is to determine if the diving profile of deep-sea diving alters the risk of developing DCS in divers. This study is a retrospective cohort study of 6,050 deep sea diving events and the corresponding 331 instances of DCS experienced by the underwater divers. The primary outcome was DCS development between the “saturation” and “repetitive and multi-level” experimental dive profile cohorts, compared to the “single-dive” control dive profile cohort. This was assessed via adjusted risk ratios. Further subgroup/confounding factor analyses were additionally performed to assess secondary outcomes of DCS development as a function of dive depth, ascent rate, dive era, and inspired gas composition. Divers and dives enrolled in this study were obtained from a study by the Naval Medical Research Center titled “The Dive Profiles and Manifestations of Decompression Sickness Cases After Air and Nitrogen-Oxygen Dives.” Dives were performed and recorded from 1940-1997. IRB approval was not required as this study utilized publicly available deidentified data. With marginal DCS events excluded due to reporting discrepancies, this study evaluated a total of 5,861 deep sea dives. Primary outcome evaluated was DCS development between the “saturation”, “repetitive and multi-level”, and “single- dive” control dive profiles evaluated by adjusted risk ratios. Secondary outcomes included DCS development as a function of dive depth, ascent rate, dive era, and inspired gas composition evaluated by adjusted risk ratios. Demographic differences were assessed via Chi-square test for categorical variables and One-way ANOVA for continuous variables. Primary and secondary outcomes were analyzed via adjusted log- binomial regression models. Primary dive profile analysis consisted of (1) “single-dive” square-profile control cohort of 3,817 dives and 189 (4.95%) instances of DCS, (2) a “repetitive and multi-level (R&M)” cohort of 1,584 dives and 67 (4.23%) instances of DCS, and (3) a “saturation” cohort of 649 dives and 75 (11.56%) instances of DCS. Single dive profile subgroups included Air Gas Inspiration: 3,139 dives and 164 (5.22%) instances of DCS and Non-Air Gas Inspiration: 678 dives and 25 (3.69%) instances of DCS. Repetitive and multi-level dive profile subgroups included Air Gas Inspiration: 849 dives and 39 (4.59%) instances of DCS and Non-Air Gas Inspiration: 735 dives and 28 (3.81%) instances of DCS. Mean(SD) dive depths (meters) for the single, R&M, and saturation cohorts were 145.2 (90.8), 91.7 (38.9), 57.2 (32.2), respectively. Mean (SD) ascent rates (meters/min) for the single, R&M, and saturation cohorts were 54.0 (114.5), 18.7 (22.4), 5.5 (12.09), respectively. Conclusions and Relevance: Primary dive profile analysis reflects that divers from the “Saturation” group were 3.65 times (95% CI: 2.63 – 5.05) more likely to experience the event then divers from the “Single: Air Gas” control group. These results were statistically significant. With regards to secondary outcomes, adjusted RRs differences in inspired gas composition do not appear to be statistically significant. Dive era, expectedly, appears to play the most significant role in determining risk of DCS, and for every 10-foot increase in depth, divers were 1.03 times (95% CI: 1.01 – 1.05) more likely to experience the event. These results were statistically significant. Ascent rate analysis was unreliable as the ascent time values in the original data are “more or less useful depending on the dive type (essentially of no use in the repetitive and multi-level dives)”.