The University of Washington is conducting a study of risk factors for decompression illness (DCI) among recreational scuba divers. With this study, we hope to contribute to the diving community and understand more about characteristics that might increase the risk of DCI.

Who is conducting the study?

The investigators are:

  • Barbara Leigh, Ph.D., MPH: Senior Scientist, School of Health Sciences, University of Washington
  • Peter Cummings, MD, MPH: Associate Professor of Epidemiology, University of Washington
  • Richard Dunford, MS: Operations director, Virginia Mason Medical Center Hyperbaric Department, Seattle

What is the purpose of the study?

In scuba diving, the major risk factors for DCI are exceeding established depth-time limits and ascending too quickly. However, the majority of DCS incidents among recreational divers (58%) are so-called "unexpected hits," occurring after a dive that was within the recommended depth and time limits.[1]

The existence of such cases, along with the variability and seeming unpredictability of DCI, has led to speculation about environmental and individual risk factors for DCI. For example, personal attributes such as age, gender, and physical fitness are suspected to affect the risk, as are some situational factors such as hydration, cold, etc. However, most of these suppositions are based on anecdote and conventional wisdom rather than scientific study.

Studies of DCI are difficult to conduct, largely because it so rarely occurs in sport diving. (The incidence of DCS in recreational divers is unknown, but is estimated to be approximately 10/100,000 dives). A second problem for research on DCI is the lack of a defined population of sport divers or their dives. The Divers Alert Network (DAN), an international organization dedicated to the safety and education of recreational divers, collects data on the characteristics of divers treated for DCI, but without information on untreated divers, we cannot tell whether certain characteristics increase the risk of DCI.

In our study, we are comparing divers treated for DCI to uninjured divers to try to identify characteristics that might affect DCI risk.

How is the study conducted?

This study uses a case-control design.[2,3,4] With this method, cases (those who develop disease or experience injury) and controls (without disease or injury) are selected, and the groups are compared on characteristics of interest. The case-control study is particularly efficient for studying rare conditions: instead of following up large numbers of people to identify sufficient numbers with the outcome of interest, the investigator identifies adequate numbers of people with and without the condition at the outset.[2,3,5] Because of its advantages, the case-control study is the most common epidemiologic design in study literature today.[2,6]

In this study, the cases are divers treated for decompression illness at the Virginia Mason Medical Center (VMMC) hyperbaric facility in Seattle. The Hyperbaric Unit treats approximately 75 cases of decompression sickness each year, placing it among the top five active treatment chambers for recreational divers in North America and the Caribbean. DAN has designated VMMC as the Pacific Northwest Regional Referral Center for diving accidents.

Divers in the control group are recruited from the sites where cases were injured. This strategy ensures that (1) cases and controls are selected from the same population, (2) controls are at risk for similar injury, (3) cases and controls are both interviewed directly after the period at risk, and (4) potential confounders of time and geography are controlled.

Cases are contacted and interviewed by telephone as soon as possible after their hyperbaric treatment. Controls are contacted and interviewed by telephone, with a delay between dive and interview comparable to that of the case diver.

The telephone interview takes about 15 minutes, and includes questions about personal attributes (age, health status, height and weight, smoking, diving experience) and dive characteristics (problems with the dive, medication use, hydration). All interview responses are completely confidential, and are used for research purposes only.

Can I find out the results?

We expect it will take several years to accumulate sufficient cases to analyze. We hope to produce a scientific paper from our findings. In the meantime, you can address questions to us at


[1] Was he within the tables? Alert Diver, Spring 1989, 5: 9.

[2] Hennekens CH, Buring JE (1987). Epidemiology in medicine. Boston/Toronto: Little, Brown.

[3] Schesselman JJ (1982). Case-control studies. New York: Oxford University Press.

[4] Rothman K (1986). Modern epidemiology. Boston/Toronto: Little, Brown & Co.

[5] Wacholder S (1995). Design issues in case-control studies. Statistical Methods in Medical Research. 4: 293 309.

[6] Cole P (1979). The evolving case-control study. Journal of Chronic Diseases. 32: 15-27.