Physician Adoption of Hyperbaric Oxygen Therapy in the Treatment of Chronic Wounds
C hronic wounds are generally defined as wounds that do not heal in a timely fashion1 or more specifically, wounds which “fail to proceed through an orderly and timely process to produce anatomic and functional integrity” after receiving appropriate medical interventions.2 Of the numerous types of chronic wounds, the most common are pressure ulcers, diabetic ulcers, and venous stasis ulcers, which together account for approximately 70% of all chronic wounds.1 Patients with chronic wounds often experience a diminishing quality of life due to anxiety and disability related to their medical condition.3 Moreover, healthcare costs associated with chronic wounds in the US are estimated to be $5 billion to $9 billion each year.2
Clinicians trying to effectively cope with chronic wounds constantly seek treatment options. Hyperbaric oxygen therapy (HBOT) is one of a number of alternatives identified over the past several years.4 This therapy has existed for many decades and has been used previously to treat other medical conditions, such as decompression sicknesses, gas embolism, gas gangrene, and carbon monoxide poisoning.5 Hyperbaric oxygen therapy to enhance wound healing involves intermittent inhalation of 100% oxygen in a chamber with elevated pressure. The underlying therapeutic mechanisms of this therapy have been shown to improve tissue oxygenation, increase fibroblast proliferation and white blood cell killing capacity during periods of hyperoxia, and stimulate angiogenesis during periods of relative hypoxia.6
Over the past decade, many prospective clinical studies have shown that HBOT significantly reduces the risk of lower extremity amputation for patients with diabetic ulcers7-11 and improves the rate of healing in the treatment of venous ulcers.12 Despite growing evidence regarding its efficacy, however, HBOT use in wound care still seems to be limited to and concentrated in only a few states within the US.13 This situation may result from the lack of robust clinical efficacy evidence secondary to the poor quality of HBOT study designs (eg, small sample size, poor documentation of wound size or severity) in most existing published clinical reports14 as well as factors such as physician unfamiliarity with HBOT or the unavailability of hyperbaric oxygen chambers.
Currently, little is known about underlying factors that are likely to influence physician adoption of HBOT. This study was designed to identify factors associated with the adoption of HBOT among physicians who specialize in wound care.
The conceptual framework for this study is based on Rogers’ diffusion of innovation model.15 This framework has been widely used to study physician adoption medical technology16,17 and to establish diffusion guidelines for medical technologies.18 According to Rogers, the individual adoption process usually involves a series of stages that include gaining knowledge about an innovation (Knowledge Stage), developing a positive or negative attitude toward that innovation (Persuasion Stage), and accepting or rejecting the innovation (Decision Stage). In the healthcare arena, this adoption process can be influenced by various factors, including physician characteristics, the innovative characteristics of a medical technology, communication channels, and the social system (see Figure 1). Utilizing Rogers’ framework, the specific objectives of this study were to: 1) understand the current knowledge of, attitudes toward, and adoption of HBOT among physicians specializing in wound care, and 2) determine whether factors representing Rogers’ framework are significantly associated with physician adoption of HBOT in wound care.
Factors Affecting Physician Technology Adoption
Physician characteristics. Based on previous studies, physician characteristics influencing physician adoption behavior with respect to medical technologies typically include personal and demographic attributes, such as age,19-23 gender,24 years of medical practice,25,26 and practice specialty.21,23,27 These factors were included in this study to assess their relevance to the adoption of HBOT. In addition, physician subjective knowledge about HBOT was examined to determine its effect on physician adoption of HBOT.
Innovative characteristics. These factors address physician perception of the relative advantages of medical technologies (eg, efficacy or cost-effectiveness).15 Physicians usually form their perceptions of medical technologies through assessment of the published clinical evidence, their own experience with technologies, and interaction with their peers.15,16 Previous studies indicate that when physicians perceive a medical technology has a relative advantage over its competitive alternatives, it is more likely or more quickly adopted.16,20,28 In the current study, the innovative characteristics addressed were wound care specialists’ perceptions of HBOT side effects and its benefits in wound healing, cost saving, and reduction of lower extremity amputations.
Communication channels. A communication channel is the means by which information about a medical technology moves from one source to another among its potential adopters.15 In this study, the authors sought to identify which communication channels physicians use frequently to stay informed about evidence regarding new wound treatments (eg, scientific medical journals, symposia, traditional and Internet-based continuing medical education programs, and contact with peers). This study examined these communication channels to determine which ones significantly affect physician knowledge of, attitudes toward, and adoption of HBOT.
Social system. A social system is a set of interrelated organizations and individuals that work together to achieve a common goal.15 In this context, the social system involves interrelationships among patients, providers, and other dynamic players and factors within the healthcare system. Based on the existing literature, factors associated with the healthcare system that often influence physician adoption of medical technologies include type of healthcare facility where physicians practice (eg, hospitals or clinics, for profit or not-for-profit status)19,23,25,27; location of the organizations19; patient volume21,23; and patient preference.16,28,29 In this study, all of these factors were examined to determine their relevancy to the adoption of HBOT in wound care. In addition, the level of local market competition and the size of the community where physicians practice also were included in the study to determine whether they influence physician adoption of HBOT.
Questionnaire development. The authors developed a questionnaire for this study that was mailed to and pretested by 10 physicians whose practice involved either wound care or hyperbaric medicine. Based on their suggestions and comments, the questionnaire was further modified. Of the 23 items in the final version, seven assessed variables associated with the social system, including patient volume, patient preference, level of competition in the local wound care market, size of the community, and type, ownership, and location of wound care facilities. Five items measured physician perceptions of HBOT in the treatment of chronic wounds (sample item: “Please indicate your level of agreement with the following statement: HBOT is an effective adjunctive treatment for chronic ischemic non-healing wounds”). Likert scale scores, ranging from 1 (strongly disagree) to 5 (strongly agree), were used to measure attitudinal factors. Internal consistency among these five items was high (Cronbach alpha = .92). Factor analysis using the principal axis extraction method also confirmed that only one factor existed and explained 72% of the variance among these items. For statistical analysis of this study, scores from these five items were compiled into one composite index and recoded into two levels (“disagree” = 5 to 17 and “agree” = 18 to 25) to represent physician attitude toward HBOT in wound care.
One item evaluated respondents’ reliance on various communication channels, such as scientific journals, peers, symposia, and continuing medical education (CME) programs, that physicians use to stay informed about wound care treatments (sample item: “How often have you depended on any of the following as a source of information to stay informed about the evidence for wound care treatments over the past 12 months?”). A 5-point Likert scale, ranging from 1 (never) to 5 (very often) was used to measure responses to this item.
Ten items gathered information about physician characteristics, such as age, gender, adoption of HBOT, physician’s subjective knowledge about HBOT, medical specialty other than wound care, and years of medical practice. It is important to note that the dependent variable of the analysis, adoption of HBOT, was determined by asking respondents, “Have you ever used HBOT to treat your patients with chronic wounds or referred your patients to other facilities for HBOT during the past 12 months?”
Four additional items measured physician subjective knowledge of HBOT (sample item: “How familiar are you with the concept of oxygen under pressure working as a drug?”). A 5-point Likert scale, with scores ranging from 1 (not a bit) to 5 (very much), was used for these items. Internal consistency among these four items was high (Cronbach alpha = .92.) Factor analysis using the principal axis extraction method confirmed that only one factor existed and explained 75% of the variance among these four items. For statistical analysis of this study, scores from these four items were compiled into one composite score and recoded into two levels (“none or little knowledge” = 4 to 10 and “some or more knowledge” = 11 to 20) to represent the level of the physician’s subjective knowledge of HBOT.
Sample and survey distribution. The questionnaire — along with a cover letter, study description, stamped return envelope, and letter of support from the Undersea and Hyperbaric Medical Society (UHMS) — was distributed on April 14, 2003 to 653 physicians who had received board certification in wound care from the American Academy of Wound Management (AAWM) before December 1, 2001. All certified physician members of the AAWM were surveyed to achieve a sufficient sample size. To increase the response rate, non-responders received a follow-up survey package after 3 weeks. Data collection for the study ended on July 1, 2003.
Statistical analysis. Descriptive statistics were used to provide an overview of the distribution of scores on each measure. In the descriptive statistics, particular emphasis was placed on assessing the current status of physician knowledge of and attitudes toward HBOT. Second, the chi-square test of independence was used to determine the bivariate association between each independent factor and the dependent variable. Finally, multiple logistic regression analysis was used to estimate the strength of association while controlling for the effects of other individual factors. The significance level for all statistical analyses was set at .05. SPSS 10.0 for Windows (SPSS Inc., Chicago, Ill.) was used to conduct all statistical analyses.
A total of 653 surveys were distributed; 44 were returned as undeliverable. Of the remaining 609 potential respondents, 260 (43%) physicians answered the study questionnaire during the data collection period. Of those, 14 were excluded from analysis because the physicians no longer practiced wound care. Respondents were unevenly distributed across 42 states in the US (see Table 1). Most practiced wound care in Florida, Pennsylvania, New York, and Texas, accounting for 109 (44%) of the respondents. No respondents practiced in Minnesota, West Virginia, Idaho, North Dakota, South Dakota, and Montana. The difference in geographic distribution of the respondents and non-respondents was examined using the chi-square test. No significant difference (P = .061) in geographic distribution was found between these two groups.
Descriptive statistics of respondent characteristics are shown in Table 2. Most (211, 86%) of the respondents were men, age 40 years or older (183, 74%), who specialized in podiatric medicine (89, 36%), practiced wound care in a community with a population size of >100,000 residents (168, 69%), and treated P = .008), medical specialty (P <.001 type="" of="" wound="" care="" facility="">P =.001), size of community (P = .002), geographic location (P <.001 patient="" volume="">P <.001 voluntary="" patient="" request="" of="" hbot="">P <.001 physician="" attitude="" toward="">P <.001 and="" physician="" knowledge="" of="">P <.001 hbot="" and="" reliance="" on="" peers="">P = .022) as a communication channel. No significant differences between groups were found with respect to physician age, years of medical practice, type of facility ownership, level of local wound care competition, and type of communication channels.
More than 60% of the respondents reported a relatively high level of familiarity with HBOT with respect to its therapeutic mechanisms, potential risks, and applications. The average scores on each knowledge item ranged from 3.71 to 4.02. When the respondents’ subjective knowledge was stratified into two groups (adopters and non-adopters), a larger percentage (71% to 84%) of the respondents in the adopter group expressed a higher level of familiarity with HBOT across the four items than the non-adopter group (32% to 47%). Results also indicate that some (16% to 29%) of the respondents who adopted HBOT still had inadequate knowledge of this therapy (see Table 3).
An assessment of respondents’ attitudes toward the effects of HBOT shows three important findings. First, a large proportion (72% to 86%) of all respondents agreed with the statement that HBOT has beneficial effects on wound healing. However, only about 48% of the respondents agreed with the statement that HBOT can reduce the treatment costs of chronic wounds. Second, among respondents who adopted HBOT, only a small percentage (2% to 4%) disagreed with the statement that HBOT has beneficial effects on wound healing. Third, more than 50% of the non-adopters agreed with all of the statements regarding the beneficial effects of HBOT on wound healing except statements concerning potential cost saving (19%) (see Table 4).
The effect of the independent variables on the adoption of HBOT (after adjusting for the effects of other variables in the regression model) is shown in Table 5. Based on the results of the multiple logistic regression analysis, five factors remained significantly associated with the adoption of HBOT:
• community size >100,000 residents (adjusted odds ratio [OR] = 2.29, 95% confidence interval [CI] = 1.05 to 5.04)
• voluntary patient request of HBOT (adjusted OR = 5.38, 95% CI = 2.50 to 11.56) – physicians whose patients requested HBOT at least once over the past 12 months were approximately six times more likely to adopt HBOT than physicians whose patients never requested this therapy
• physician attitude toward HBOT (adjusted OR = 3.38, 95% CI = 1.49 to 7.66), suggesting that respondents with a favorable attitude toward HBOT were more likely to adopt this therapy than physicians with an unfavorable attitude
• physician knowledge of HBOT (adjusted OR = 5.33, 95% CI = 1.72 to 16.49) – respondents who had adequate knowledge of HBOT were much more likely to adopt it than respondents who had little or no knowledge of this therapy
• location (Florida and Texas) where respondents practiced (adjusted OR = 3.44, 95% CI = 1.24 to 9.54), suggesting that respondents who practiced in either Florida or Texas had a higher likelihood of adopting HBOT than respondents who practiced in other states.
Other variables (eg, gender, type of wound care facility, and size of community) that previously were found to be significantly associated with the adoption of HBOT during the bivariate analyses were no longer significant after adjusting for the effects of other factors in the logistic regression model.
One of the goals of this study was to understand the current adoption, knowledge of, and attitudes toward HBOT among physicians specializing in wound care. The results of this study indicate that the majority (68%) of physician respondents already have used HBOT to treat their patients or referred their patients to other facilities for HBOT during the 12 months preceding the study (32% have not used HBOT in their practice). Among the adopters, 35% practiced in Florida and Texas while the remaining 65% were spread out across 40 different states. The rate of respondents adopting HBOT before the study period was 83% and 100% in Florida and Texas, respectively. The adoption rate of respondents in other major states (New York, Pennsylvania, and California) was 57%, 36%, and 57%, respectively. These rates suggest that HBOT has been adopted by the majority of currently practicing AAWM board-certified wound care physicians but the therapy seems widely accepted in only a few states (Florida and Texas). Although this finding is somewhat preliminary due to potential selection bias, it improves understanding of the current diffusion of HBOT in the US among physicians specializing in wound care.
Gaining knowledge about an innovation usually is the initial step in the individual-level adoption process. An individual with more knowledge about an innovation is more likely to adopt it.1 This study also evaluated current physician knowledge of HBOT. Based on the study’s results, most (60%) of the respondents reported adequate knowledge of HBOT regarding its therapeutic mechanisms, applications, and potential risks. This high percentage corresponds to the high level of adoption of HBOT among the respondents. It is important to note that the survey of physician knowledge of HBOT in this study was a subjective self-assessment and not validated by ascertaining exact physician knowledge or skill. Nonetheless, nearly 40% of respondents had only “some” or “little knowledge” about HBOT. Among those respondents who had less knowledge about HBOT, most (88%) were podiatrists. Thus, if anyone in the wound care industry wants to significantly increase the use of HBOT, educational or promotional efforts should be targeted to podiatrists.
Developing attitudes toward an innovation is usually the second step in the adoption process. Because a person with a favorable attitude toward an innovation is more likely to adopt it,15 physician attitudes toward HBOT in the treatment of chronic wounds were assessed. Two inferences can be drawn from the study’s findings. First, the majority of respondents strongly agreed with the various beneficial effects of HBOT in the treatment of chronic wounds, with adopters having a more favorable attitude toward HBOT than non-adopters. Despite lack of strong evidence from well-designed clinical studies as reported by Wunderlich et al,14 this finding indicates that most wound care specialists agree with the survey statement that HBOT has beneficial effects on chronic ischemic wounds. Second, considerable disagreement was noted regarding the statement about HBOT cost-effectiveness in the treatment of chronic wounds. Only 48% of the respondents believed HBOT is cost-effective. This disagreement may result from insufficient evidence of the cost-effectiveness of HBOT in the treatment of chronic wounds.30 Since the issue of cost will become increasingly important as stakeholders, especially payors and providers, become more cost conscious, HBOT cost-effectiveness studies are needed.
Another purpose of this study was to examine additional variables in Rogers’ diffusion of innovation framework that can potentially influence physician adoption of HBOT. In the bivariate analysis, the adoption of HBOT was significantly associated with multiple factors, including physician gender (crude OR = 2.60, 95% CI = 1.26 to 5.39), podiatrist certification (crude OR = 0.30, 95% CI = 0.16 to 0.56), hospital-based wound care facilities (crude OR =2.49, 95% CI = 1.41 to 4.39), size of community >100,000 residents (crude OR = 2.48, 95% CI = 1.41 to 4.38), patient volume >30 per week (crude OR = 4.25, 95% CI = 1.98 to 9.10), voluntary patient request of HBOT (crude OR = 9.48, 95% CI = 5.13 to 17.50), location (crude OR = 4.72, 95% CI = 2.13 to 10.48), frequent use of peers as a information source (crude OR = 1.88, 95% CI = 1.09 to 3.22), and physician knowledge of (crude OR = 13.93, 95% CI = 5.45 to 35.60) and attitude toward (crude OR = 4.17, 95% CI = 2.29 to 7.59) HBOT. Of these factors, only five remained significant after adjusting for the effects of the other factors in the multiple logistic regression model: community size >100,000 residents, voluntary patient request of HBOT, physician knowledge of and attitude toward HBOT, and practice location (see Table 5).
The significant association between community size and adoption of HBOT may result from the fact that most hyperbaric oxygen chambers are located in urban areas.31 Hence, physicians practicing in larger communities normally have a greater chance of using HBOT. This study’s findings confirm previous reports in which patient preference was demonstrated to have a positive influence on physician adoption of medical technologies.16,28,29 Physicians in this study whose patients requested provision of HBOT at least once were more likely to adopt it than physicians who never had requests for this therapy. Such a strong association may imply that educating a select group of patients or lay population to increase service demand may be an effective way to rapidly influence physicians to adopt the therapy.
Physician knowledge of HBOT was another important factor that influenced its adoption. Physicians more familiar with HBOT were more likely than those less familiar with HBOT to use this therapy. Furthermore, level of physician HBOT knowledge was significantly associated with frequent reliance for information on peers (crude OR = 4.62, 95% CI = 2.04 to 10.43) and symposia (crude OR = 2.44, 95% CI = 1.16 to 5.15). These important information sources underscore the need for identifying and sponsoring local physician champions and supporting related clinical symposia — useful strategies to effectively disseminate information regarding HBOT among physicians.
Individual items within the knowledge measure (see Table 3) were examined using single logistic regression to determine which item(s) had the strongest effect on the adoption of HBOT. The regression analysis shows that physician familiarity with the therapeutic mechanisms of HBOT in wound healing had the greatest impact (crude OR = 17.76, 95% CI = 3.93 to 80.37) on the adoption of HBOT among all four items. A possible explanation for this finding is that an individual is more likely to adopt new technology that has relevance to his/her specific situation.15 Because the study population was comprised of physicians specializing, and board certified, in wound care, knowing more about the therapeutic mechanisms of HBOT specific to wound healing would increase the likelihood of adopting HBOT in practice. The message used to promote HBOT in a specific medical specialty should be tailored to its members’ interests.
Multiple logistic regression results indicate that physician attitude toward the beneficial effects of HBOT is significantly associated with physician adoption of this therapy. Respondents with a favorable attitude were more likely than those who had an unfavorable attitude to adopt this therapy. When items within the attitudinal measure (see Table 4) were examined to determine which had the strongest effect on the HBOT adoption, physician attitude toward the cost-effectiveness of therapy had the largest effect (crude OR = 6.50, 95% CI= 3.46 to 12.23) among all five items. This is probably because HBOT is expensive in the US and considered an adjunct treatment. As a result, a physician’s decision to use HBOT is likely to be based on whether the potential benefits outweigh its costs. This finding suggests that information or evidence used to promote HBOT use in the medical community should include evidence of cost-effectiveness.
Practice location was significantly associated with the adoption of HBOT. Physicians practicing wound care in Florida and Texas were significantly more likely to use HBOT than those who practice in other states. No difference in physician knowledge (P = 0.072) and attitude (P = 0.101) across locations (Florida and Texas versus other states) was found based on the chi-square test. However, Florida and Texas have more HBOT treatment centers. According to the 2001 Chamber Registry of the Undersea and Hyperbaric Medical Society (UHMS),31 110 (29%) of the 377 registered HBOT treatment centers in the US are in Florida and Texas. Given the greater availability, physicians in these states may be more likely to use HBOT to treat their patients or refer their patients to other local facilities for HBOT. Another factor potentially contributing to the higher adoption of HBOT in both states may be the higher level of local wound care competition perceived by the respondents. In this study, respondents reported that the mean level of local wound care competition in Florida and Texas was significantly higher (P
Several limitations of this study need to be recognized in order to properly interpret the study findings. First, generalizing the findings of the study to all physicians specializing in wound care may be problematic because more than 50% of the potential study population did not participate and not all wound care physicians are certified by the AAWM. In addition, sampling bias may have resulted in a sample that is particularly interested in HBOT and, therefore, more likely to adopt it or report more knowledge and favorable attitudes. Based on the available information, respondents were no different from non-respondents with regard to geographic location. Second, because this is a cross-sectional study in which self-reported data were collected at only one point in time, the causality of relationships between the dependent and independent variables is difficult to determine. Longitudinal studies of physician adoption of HBOT would provide useful information regarding the dynamics of the adoption process.
Modern HBOT technology has existed for more than five decades and has been approved by the US Food and Drug Administration for the treatment of many medical conditions, including chronic wounds. Growing evidence of its efficacy in wound healing has increased the adoption of HBOT among physicians practicing in wound care. However, the increase seems to have been limited by several factors. The study’s findings offer useful insights to those interested in promoting the use of HBOT in the medical community. First, educating patients or the lay population about the benefits of HBOT will likely influence physicians to adopt it. Second, local physician champions and symposia are effective communication channels that may facilitate the adoption of HBOT in a specific geographic area or medical specialty. Finally, when promoting its uses, focusing on the cost-effectiveness in addition to the therapeutic effectiveness of HBOT will likely have a positive influence on physician adoption of this therapy.