What is the risk of decompression illness?

The short answer

The risk of decompression illness among recreational divers is approximately 10-20 hits per 100,000 dives.

The long answer

The rate of a disease or disorder is defined as (the number of new cases of the disorder during a given period of time)/(the total population at risk). For example, the cancer mortality rate for 1980 in the U.S. was 183.8/100,000/year. So, to calculate the rate of decompression illness, we need to know the number of cases of DCI during a certain period of time and the number of dives that gave rise to those cases.

In Navy and commercial diving, because records are kept for each dive, DCI rates can be calculated. For example, in 706,259 Navy dives from 1968 to 1981, there were 426 cases of decompression sickness and 37 cases of air embolism; the rate of DCI (DCS +AGE) was thus 65/100,000 dives.[1] A similar analysis of Navy dives from 1960 to 1969 yielded a rate of 47/100,000 dives[2]. In commercial diving, the rate of DCI varies greatly by location and procedures (for example, procedures are different for U.K. companies working in the North Sea than for U.S. companies working in U.S. waters). For North Sea operations, estimates range from 485/100,000 dives in the early 1980s [3,4] to 70/100,000 in 1990. [5]

In recreational diving, although there are some statistics kept on the number of cases of DCI occurring each year, there are usually no good estimates of the total number of dives undertaken. There are a few studies, however, that have gathered information on DCI cases and on the underlying number of dives.

Field studies

Gilliam [6,7] summarized statistics from a dive operation aboard the Ocean Spirit, a 457-foot cruise ship that offered diving to 160 divers on each one-week cruise in the Caribbean. From March 1989 to March 1990, 77,680 dives were completed by customers and staff. Eight cases of DCS and two cases of AGE occurred during this period, and all cases were treated in the ship's on-board recompression chamber. Seven of the 8 divers with DCS were using tables (five of these were diving within table limits); the single computer diver with DCS had violated the computer's ceiling. The rate of DCI (DCS + AGE) was 12.9/100,000 dives, and the rate of DCS alone was 10.3/100,000 dives. The rate of so-called "unexpected" DCS (DCS that occurred without violating time-depth limits) was 6.4/100,000 dives.

In a study of sport diving in the U.S. military community on Okinawa, Arness [8] used the number of scuba tanks filled in U.S. facilities to estimate the number of dives occurring in this population. Because equipment rental and air fills are very expensive at commercial Japanese operations, virtually all tank fills come from the U.S. facility. The military community is medically isolated and self-sufficient: all diving accidents among military personnel are treated at the military facility. Over seven years, 94 cases of DCI were treated, with an estimated 100,000 dives occurring yearly in this population. The rate of DCI was 13.4/100,000 dives, and the rate of "unexpected" DCI (DCI that occurred with no time-depth violations or missed decompression stops) was 10/100,000 dives.

A similar study in British Columbia, Canada [20] used the number of tank fills to estimate the number of dives conducted. For 14 months, dive shops and charter operators in the province were asked to count the number of tanks that they filled for recreational diving. During this period, 14 DCI injuries were treated in local hyperbaric chambers, with an estimated 146,291 dives occurring. The rate of DCI was 9.57/100,000. However, because only 65% of dive shops and dive operators participated, the number of tank fills underestimates the actual number of dives. Therefore, the true DCI rate is lower than this estimation, by an unknown amount.

Hart et al. [9] reported on diving accidents at a large inland diving center in England. The number of divers visiting the center each year was obtained from the dive center's database. From 1992 to 1996, there were 25 diving accidents at the center, seven of which resulted in death. The rate of injuries varied by year, from 4/100,000 divers in 1992 to 15.4/100,000 divers in 1996; over the five years, the rate of injuries was 10.5/100,000 divers. Note two features of this study that make it difficult to estimate DCI rates: 1) Accidents included not only DCI, but other injuries including barotrauma and drowning. 2) The denominator is the number of divers, rather than the number of dives, and it is not clear whether this number includes repeat visitors.

To investigate DCI rates at Scapa Flow, off the northeast coast of Scotland, Trevett [21] used records supplied by the operators of dive charters in 1999 and 2000. The total number of dives conducted was estimated by multiplying the number of boat diver-days by two (to account for the customary two-tank day trip) and adding an estimated number of shore dives provided by local dive schools (the vast majority of diving in this area is by boat). The estimated rate of DCI was 20/100,000 in 1999 and 46/100,000 in 2000. A second estimate used a different method to calculate the total number of dives conducted: a random sample of 258 divers logged all of their dives during the 2002 season, and the total number of annual dives was estimated by multiplying the average number of dives (9.6 dives per diver) by the total number of visiting divers. The resulting DCI estimate was somewhat larger: 87.4/100,000 for 2000-2002.[22]

DAN's Project Dive Exploration (PDE) is a field study that collects information on dive profiles with depth-time recorders worn by divers who then send their profiles to DAN (see DAN's description for more information on the study). The most recent report from this study [25] reports on data from 1998 to 2002 and includes information from 48,548 dives with 26 DCS cases for an overall rate of 53/100,000 dives. This rate differed by geographical area and diving style, with a rate of 10/100,000 for Caribbean liveaboard divers and 280/100,000 for divers in Scapa Flow.

Questionnaire studies

A study from DAN Europe[10,11] used data from questionnaires completed by members. Divers who enroll or renew their membership in DAN Europe are asked to complete a questionnaire that includes a question about the number of dives completed per year. From 1989 to 1993, 17,700 questionnaires were collected; divers averaged 25 dives per year for a total of 442,500 person-dives. During this period, 67 decompression accidents occurred in DAN Europe members, for a rate of 15/100,000 dives.

A survey of DAN members in the U.S. [18] yielded 1628 questionnaires that contained information on diving activity and DCI in the previous 5 years. The total number of dives was estimated from the questionnaire reports. There were 67 cases of DCI (only 30 were treated) in 198,167 dives. Assuming that all 67 cases were truly DCI gives a rate of 33.8/100,000 dives; assuming that untreated cases were not DCI gives a rate of 15.1/100,000.

Hahn [12] distributed questionnaires to members of the German Sport Divers Association in 1989 asking how many dives the members had completed in that year. The total number of dives completed by the entire membership of the Association was estimated by multiplying the average number of dives for 1989 reported in the questionnaire (17.4) by the number of club members (30,817). Ten cases of DCS were reported to insurance that year for a rate of 1.9/100,000 dives.

Wilmshurst [13] sent questionnaires to 10% of the membership of the British Sub-Aqua Club (BSAC) in 1976, 1980, and 1986. The total number of dives completed by BSAC members was estimated by multiplying the average number of yearly dives reported on the questionnaire by the number of BSAC members. In 1986, 52 cases of DCI (DCS or AGE) were reported to the BSAC for a rate of 4.69/100,000 dives.

Bangasser [14] sent questionnaires to male and female dive instructors, asking for the number of dives they had completed and whether they had ever had DCS. There were 13 cases of confirmed or suspected DCS in 87,280 reported dives, for a rate of 14.9/100,000 dives.

St. Leger Dowse et al.[15] distributed questionnaires to divers through British sport diving associations, dive shops, and the BSAC. Divers were asked to note the total number of dives they had completed and whether they had ever had DCI. Out of a total of 458,827 dives reported, there were 87 diagnosed cases of DCI and 72 self-diagnosed cases. The rate of diagnosed DCI was 19/100,000 dives; the rate of diagnosed and self-diagnosed DCI was 34.6/100,000 dives.

Nakayama et al. [19] administered questionnaires over a 6-year period at a popular diving location in Japan. 60 DCI incidents had been experienced by the 2975 divers who completed the questionnaire, and these divers reported 1.14 million dives. The rate of DCI was 5.3/100,000 dives.

Hagberg [23] sent questionnaires to divemasters and instructors listed with PADI, NAUI and CMAS in Sweden (2380 divers), asking for information about diving and diving injuries that occurred in 1999. The 1737 divers who responded reported performing 127,256 dives during 1999, and 190 divers reported experiencing DCS symptoms during that same year. The rate of symptoms was 149/100,000 dives, which is quite a bit higher than estimates from other questionnaire studies. However, as the authors note, many DCS symptoms go untreated, and the rate of treated DCS would be much lower. Although information about treatment was not collected in this study, there were only 23 DCS cases treated in Sweden in 1999.

Taylor [24] contacted 29 dive clubs in Australia and 28 clubs in the U.S., and arranged for club representatives to distribute questionnaires at club meetings. The 709 participating divers reported 182,951 dives with 31 cases of DCS, for a rate of 17/100,000 dives.


Because it is difficult to study a defined population of recreational divers and their dives, the exact rate of DCI is difficult to calculate. Two studies of relatively well-defined populations in tropical waters [6,8] reported similar estimates of 12.9 and 13.4 hits per 100,000 dives, while a study in a cold-water area with deep, square-profile wreck diving estimated a higher rate of 20 to 80 hits per 100,000 dives [21,22]. Using questionnaires of sport divers' self-reports, other studies have estimated the DCI rate as low as 1.9/100,000 dives and as high as 19/100,000 dives.

Note that because DCI is rare, the calculated estimates can be unreliable. For example, DAN's calculated rate of DCI for day boats and shore dives in the Caribbean is based on only 5 treated DCI cases (as of their 2004 report), and the reported rate for these dives has fluctuated from 32 to 86/100,000 dives over several years of reporting.

Because the rate of DCI is a fraction, error in the estimates can be introduced by errors in either the numerator or denominator. For example, if the number of DCI cases is underestimated, then the estimated rate is lower than the true rate; if the number of total dives is underestimated, the estimated rate is higher than the true rate. Some sources of error include:

  • Underreporting of DCI: There are several reasons for DCI symptoms to go untreated, including the innocuous appearance of some DCS symptoms and the stigma attached to DCI among divers. The proportion of DCI symptoms that go untreated is unknown, but DAN estimates it may be as high as 50%. If the number of DCI cases is underestimated, then the estimated rate of DCI is also underestimated.
  • Errors in diagnosing DCI: There is no definitive diagnosis of DCS; instead, determination of DCS depends on assessment of several factors, including recent diving activity, characteristic symptoms, and improvement of symptoms with recompression[16,17]. Other conditions can be mistaken for decompression illness; for example, in Arness' study of diving on Okinawa, ten of 104 treated cases were assigned another diagnosis (including musculoskeletal injury, gastroenteritis, and barotrauma).
  • Errors in estimates of the population of dives: Particularly in questionnaire studies, the number of estimated total dives could be inaccurate. The estimates in questionnaire studies come from those divers who filled out the questionnaire, who may not represent the average diver. The true number of dives might be either more or less than the estimated number of dives.

Although these errors can lead to underestimates or overestimates of risk, it is clear that the risk of DCI in sport diving is low. However, knowing that the risk is low is of no comfort if you get bent. You can minimize your risk by doing shallower dives (DCI risk is lower for dives shallower than 100 feet), ascending slowly, and performing deep stops (if warranted) or safety stops.


[1] Blood C, Hoiberg A. Analyses of variables underlying U.S. Navy diving accidents. Undersea Biomedical Research 1985;12:351-360.

[2] Biersner RJ. Factors in 171 navy diving decompression accidents occurring between 1960-1969. Aviation Space and Environmental Medicine 1975;46:1069-73.

[3] Giles R. (1989). Decompression sickness from commercial offshore air-diving operations on the UK continental shelf during 1982 to 1988. Report from the UK Department of Energy.

[4] Overland T. Oceaneering International. In: Lang MA, Vann RD, eds. Proceedings of the Repetitive Diving Workshop. Costa Mesa, CA: American Academy of Underwater Sciences,1992:89-101.

[5] Laden G. Incidence of decompression sickness arising from air diving operations (letter). Undersea and Hyperbaric Medicine 1998; 25(4):237-9.

[6] Gilliam B. Evaluation of DCS incidence of multi-day repetitive diving for 77,680 sports dives. In: Lang MA, Vann RD, eds. Proceedings of the Repetitive Diving Workshop. Costa Mesa, CA: American Academy of Underwater Sciences,1992:15-24.

[7] Hill RK, Gilliam BC. A comparison of the incidence of decompression sickness in men and women divers using decompression tables and diving computers for 77,680 dives. In: Jaap W, ed. Diving for Science. 1990. American Academy of Underwater Sciences, Costa Mesa, CA. 1990:157-162.

[8] Arness MK. Scuba decompression illness and diving fatalities in an overseas military community. Aviation Space and Environmental Medicine 1997; 68:325-33.

[9] Hart AJ, White SA, Conboy PJ, Bodiwala G, Quinton D. Open water scuba diving accidents at Leicester: five years' experience. J Accid Emerg Med 1999;16:198-200.

[10] Marroni A. "DAN Europe President Dr. Marroni talks about DAN Europe's role in reporting dive injuries." Alert Diver, January/February 1999:6-7.

[11] Marroni A. Recreational diving today, risk evaluation and problem management. In: Cimsit M, ed. Proceedings of the 20th Annual Meeting of the EUBS. Istanbul: 1994:121-134. Summarized in DAN Europe Project Safe Dive RFO Manual.

[12] Hahn MH. 1989 VDST statistics. In: Lang MA, Vann RD, eds. Proceedings of the Repetitive Diving Workshop. Costa Mesa, CA: American Academy of Underwater Sciences, 1992:13-14.

[13] Wilmshurst P. Analysis of decompression accidents in amateur divers. Progress in Underwater Science 1990; 15:31-37

[14] Bangasser S. Medical profile of the woman scuba diver. In: NAUI Proceedings of the Tenth International Conference on Underwater Education, 1980:12-13.

[15] St. Leger Dowse M, Bryson, P, Gunby A, Fife W. Men and Women in Diving. Diving Diseases Research Center, Plymouth, UK: 1994.

[16] Moon RE, Gorman DF (1993). Treatment of the decompression disorders. In: Bennett PB, Elliott DH, eds. The physiology and medicine of diving (pp. 506-541). London: W.B. Saunders Company.

[17] Moon RE (1997). Treatment of decompression sickness and arterial gas embolism. In: Bove AA, ed. Diving medicine. Philadelphia: W.B. Saunders Company.

[18] Dear, G., Uguccioni, D.M., Dovenbarger, J.A., Thalman, E.D., Cudahy, E., & Hanson, E. (1999). Estimated DCI incidence in a select group of recreational divers. Annual Meeting of the Undersea and Hyperbaric Medical Society.

[19] Nakayama, H., Shibayama, M., Yamami, N., Togawa, S. Takahashi, M., Mano, Y. (2003). Decompression sickness and recreational scuba divers. Emergency Medical Journal, 20: 332-334.

[20] Ladd, G., Stepan, V., and Stevens, L. (2002) New method for counting recreational scuba dives establishes risk of death and DCI injury. SPUMS Journal, 30(3):124-128.

[21] Trevett, A.J., Forbes, R.F., Rae, C.K., Sheehan, C., Ross, J. (2001) The incidence of diving accidents in sports divers in Orkney, Scotland. Presented at the Annual Meeting of the Undersea and Hyperbaric Medical Society.

[22] Trevett, A.J., Forbes, R.F., Rae, C.K., Sheehan, C. (2003). The incidence of diving accidents in recreational divers in Orkney waters--Three years experience. Presented at the Annual Meeting of the Undersea and Hyperbaric Medical Society.

[23] Hagberg, M., & Ornhagen, H. (2003) Incidence and risk factors for symptoms of decompression sickness among male and female dive masters and instructors - a retrospective cohort study. Undersea and Hyperbaric Medicine, 30, 93-102.

[24] Taylor, D.M., O'Toole, K.S., Ryan, C.M. (2003). Experienced scuba divers in Australia and the United States suffer considerable injury and morbidity. Wilderness and Environmental Medicine, 14, 83-88.

[25] Vann, R.D., Denoble, P.J., Uguccioni, D.M., Freiberger, J.J., Forbes, R., Pieper, C.F. (2004) Incidence of decompression sickness (DCS) in four recreational diving populations. Presented at the Annual Meeting of the Undersea and Hyperbaric Medical Society, Sydney, Australia (Abstract appears in Undersea and Hyperbaric Medicine 2004, 31(3), 304-305).