Download Free Cox Proportional Hazards Models For Modeling The Time To Onset Of Decompression Sickness In Hypobaric Environments Book in PDF and EPUB Free Download. You can read online Cox Proportional Hazards Models For Modeling The Time To Onset Of Decompression Sickness In Hypobaric Environments and write the review.

In this paper we fit Cox proportional hazards models to a subset of data from the Hypobaric Decompression Sickness Databank. The data bank contains records on the time to decompression sickness (DCS) and venous gas emboli (VGE) for over 130,000 person-exposures to high altitude in chamber tests. The subset we use contains 1,321 records, with 87% censoring, and has the most recent experimental tests on DCS made available from Johnson Space Center. We build on previous analyses of this data set by considering more expanded models and more detailed model assessments specific to the Cox model. Our model - which is stratified on the quartiles of the final ambient pressure at altitude - includes the final ambient pressure at altitude as a nonlinear continuous predictor, the computed tissue partial pressure of nitrogen at altitude, and whether exercise was done at altitude. We conduct various assessments of our model, many of which are recently developed in the statistical literature, and conclude where the model needs improvement. We consider the addition of frailties to the stratified Cox model, but found that no significant gain was attained above a model that does not include frailties. Finally, we validate some of the models that we fit.Thompson, Laura A. and Chhikara, Raj S. and Conkin, JohnnyJohnson Space CenterDECOMPRESSION SICKNESS; EXTRAVEHICULAR ACTIVITY; DATA BASES; PARTIAL PRESSURE; HIGH ALTITUDE TESTS; HAZARDS; QUARTILES; PRESSURE; PHYSICAL EXERCISE; NONLINEARITY; NITROGEN; EXPOSURE; ALTITUDE SIMULATION; AEROEMBOLISM
Decompression sickness (DCS) is a complex, multivariable problem. A mathematical description or model of the likelihood of DCS requires a large amount of quality research data, ideas on how to define a decompression dose using physical and physiological variables, and an appropriate analytical approach. It also requires a high-performance computer with specialized software. I have used published DCS data to develop my decompression doses, which are variants of equilibrium expressions for evolved gas.
It is important to understand the risk of serious hypobaric decompression sickness (DCS) to develop procedures and treatment responses to mitigate the risk. Since it is not ethical to conduct prospective tests about serious DCS with humans, the necessary information was gathered from 73 published reports. We hypothesize that a 4-hr 100% oxygen (O2) prebreathe results in a very low risk of serious DCS, and test this through analysis. We evaluated 258 tests containing information from 79,366 exposures in altitude chambers. Serious DCS was documented in 918 men during the tests. A risk function analysis with maximum likelihood optimization was performed to identify significant explanatory variables, and to create a predictive model for the probability of serious DCS [P(serious DCS)]. Useful variables were Tissue Ratio, the planned time spent at altitude (Talt), and whether or not repetitive exercise was performed at altitude. Tissue Ratio is P1N2/P2, where P1N2 is calculated (N2) pressure in a compartment with a 180-min half-time for N2 pressure just before ascent, and P2 is ambient pressure after ascent. A prebreathe and decompression profile Shuttle astronauts use for extravehicular activity (EVA) includes a 4-hr prebreathe with 100% O2, an ascent to P2=4.3 lb per sq. in. absolute, and a Talt=6 hr. The P(serious DCS) is: 0.0014 (0.00096-0.00196, 95% confidence interval) with exercise and 0.00025 (0.00016-0.00035) without exercise. Given 100 Shuttle EVAs to date and no report of serious DCS, the true risk is less than 0.03 with 95% confidence (Binomial Theorem). It is problematic to estimate the risk of serious DCS since it appears infrequently, even if the estimate is based on thousands of altitude chamber exposures. The true risk to astronauts may lie between the extremes of the confidence intervals since the contribution of other factors, particularly exercise, to the risk of serious DCS during EVA is unknown. A simple model that only accounts for four important variables in retrospective data is still helpful to increase our understanding about the risk of serious DCS.
Decompression sickness (DCS) is a complex, multivariable problem. A mathematical description or model of the likelihood of DCS requires a large amount of quality research data, ideas on how to define a decompression dose using physical and physiological variables, and an appropriate analytical approach. It also requires a high-performance computer with specialized software. I have used published DCS data to develop my decompression doses, which are variants of equilibrium expressions for evolved gas plus other explanatory variables. My analytical approach is survival analysis, where the time of DCS occurrence is modeled. My conclusions can be applied to simple hypobaric decompressions - ascents lasting from 5 to 30 minutes - and, after minutes to hours, to denitrogenation (prebreathing). They are also applicable to long or short exposures, and can be used whether the sufferer of DCS is at rest or exercising at altitude. Ultimately I would like my models to be applied to astronauts to reduce the risk of DCS during spacewalks, as well as to future spaceflight crews on the Moon and Mars.
Decompression sickness (DCS) is a complex, multivariable problem. A mathematical description or model of the likelihood of DCS requires a large amount of quality research data, ideas on how to define a decompression dose using physical and physiological variables, and an appropriate analytical approach. It also requires a high-performance computer with specialized software. I have used published DCS data to develop my decompression doses, which are variants of equilibrium expressions for evolved gas plus other explanatory variables. My analytical approach is survival analysis, where the time of DCS occurrence is modeled. My conclusions can be applied to simple hypobaric decompressions - ascents lasting from 5 to 30 minutes - and, after minutes to hours, to denitrogenation (prebreathing). They are also applicable to long or short exposures, and can be used whether the sufferer of DCS is at rest or exercising at altitude. Ultimately I would like my models to be applied to astronauts to reduce the risk of DCS during spacewalks, as well as to future spaceflight crews on the Moon and Mars. Conkin, Johnny Johnson Space Center NASA/TP-2001-210775, S-885, NAS 1.60:210775, JSC-CN-7170
This book is for statistical practitioners, particularly those who design and analyze studies for survival and event history data. Building on recent developments motivated by counting process and martingale theory, it shows the reader how to extend the Cox model to analyze multiple/correlated event data using marginal and random effects. The focus is on actual data examples, the analysis and interpretation of results, and computation. The book shows how these new methods can be implemented in SAS and S-Plus, including computer code, worked examples, and data sets.