5

Outcomes of the Human Intervention and Motivational Study and the Flight Attendant Drug and Alcohol Program: Analysis of the Available Evidence

The study committee aimed to ascertain characteristics of the flight attendants and pilots who took part in the Flight Attendant Drug and Alcohol Program (FADAP) and Human Intervention and Motivational Study (HIMS), respectively; the details of their treatment (e.g., length of treatment, whether or not medication-assisted treatment was offered); and measures of program satisfaction and effectiveness. To do this, the committee requested access to the FADAP and HIMS outcome databases for analysis by an independent consultant. In addition, to obtain flight attendants’ and pilots’ perspectives on the FADAP and HIMS, the committee developed a tool for anonymously gathering experiences (through a “Call for Perspectives”) and commissioned a qualitative study involving interviews with flight attendants and pilots. This chapter summarizes the findings from these efforts.

Before presenting the findings, a few comments are warranted. First, the data are characterized by low response rates and/or selection bias, which limit the conclusions we can confidently draw from them. Second, however, this chapter presents these data because it is important to describe what is and is not known. For example, the substantial amount of missing data due to low survey response rates are itself an important finding. Third, much of the data is based on self-reports. Although self-reports can be reliable and valid, they are subject to response biases, such as the social desirability bias—a tendency to answer questions in a way that will be judged favorably by others (Althubaiti, 2016). It is important to be aware of the possibility of under-reporting of substance use. For example, data available to the committee suggest that among both flight attendants and pilots, opioid misuse is relatively infrequent compared to

misuse of other substances. It is possible that the available data under-count the actual prevalence of opioid misuse in the aviation industry and continued vigilance related to opioid (mis)use is warranted.

FINDINGS FOR FADAP

Findings from an Independent Analysis of the FADAP Database

FADAP collects data on flight attendant outcomes among those who have received treatment from FADAP-approved residential treatment centers. Flight attendants are asked to complete three surveys:

1. An initial survey administered approximately three weeks into treatment (called “Initial Self-Report Survey,” first launched in 2014), which assesses the history of prior treatment, disciplinary actions at work, job performance, treatment engagement, and satisfaction with FADAP;
2. A post-treatment survey administered at three to four weeks after treatment completion (called “Post-treatment Survey,” first launched in calendar year 2020), which assesses current employment status and satisfaction with the residential treatment program; and
3. A post-treatment survey administered at one year after treatment completion (called “Follow-up Self-Report Survey,” first launched in 2014), which assesses current self-described “recovery” status, current employment status, disciplinary actions at work, current and past job performance, current and past engagement in treatment, and satisfaction with FADAP.

Flight attendants’ treatment providers are also asked to complete a survey on each flight attendant’s treatment (called “Primary Treatment Summary” survey), which assesses a flight attendant’s admission and discharge dates, whether the flight attendant completed the full course of treatment, the flight attendant’s presenting problem (e.g., alcohol, drugs, mental health), whether social supports participated in treatment, details of medication-assisted therapy (e.g., whether it was offered, accepted), and the flight attendant’s level of engagement in treatment.

The study committee contracted with Cara M. Nordberg (MPH) to conduct an independent analysis of the FADAP database. Key results are briefly summarized here, and the full report can be found in the “Resources” section of this report’s website. The data were received in November 2022. The analyzed database comprised data from 1,196 unique flight attendants who were admitted for treatment between June 2014 and October 2022

(with the latest discharge date in November 2022), and 1,172 unique treatment episodes. To be counted as a unique treatment episode by the analyst, treatment providers must have returned the “Primary Treatment Summary” survey with non-missing admission and discharge dates, and repeat treatment episodes had to be separated by at least 28 days.

Completeness of the FADAP Database

Survey response rates are important because they affect interpretation of the findings. When survey response rates are low, findings may be biased because the responses represent only a small subset of the population of interest, and respondents often differ from non-respondents in important ways. For example, prior research has shown that respondents tend to be healthier than nonrespondents (e.g., have less severe substance use and mental health problems), which means that findings based on a small subset of respondents are often biased toward healthier respondents. When survey response rates are high, there is less concern about biased findings, since most of the population (i.e., both severe and non-severe cases) is represented in the data.

Response rates varied across the FADAP surveys. The 1,196 flight attendants showed a fairly high response rate to the “Initial Self-Report Survey” (73.6%, N = 880). Among flight attendants far enough out from treatment and therefore eligible to complete the 3–4 week “Post-treatment Survey” (N = 377) and the one-year “Follow-up Self-Report Survey” (N = 835), response rates were low, at 23.3 percent (N = 88) and 29.6 percent (N = 247), respectively. The low response rate to the “Follow-up Self-Report Survey” administered at one year after treatment is noteworthy, because this is the survey that collects much of the treatment outcome data (e.g., recovery, return to work, post-treatment satisfaction with FADAP). The treatment provider response rate to the “Primary Treatment Summary” survey was high, with treatment summary data available for 82.8 percent (N = 990) of flight attendants.

To ascertain whether there were differences between respondents and nonrespondents, the analyst compared flight attendants and treatment episodes (since flight attendants could have had multiple treatment episodes) that were lost to follow-up versus those that were not. Loss to follow-up was defined as nonresponse to the one-year “Follow-up Self-Report Survey” for flight attendants/treatment episodes at least one year out from treatment (i.e., those who had completed treatment by May 31, 2021). Overall, 797 flight attendants were eligible for the one-year “Follow-up Self-Report Survey,” and 69.0 percent (N = 550) of these flight attendants were lost to follow-up (see Table 5-1). Flight attendants who were lost to follow-up did not differ from flight attendants who were not lost to follow-up in terms

TABLE 5-1 A Comparison of Flight Attendants and Treatment Episodes That Were Lost Versus Not Lost to Follow-Up

NOTES: ^a797 of the 1,196 unique flight attendants in the database were eligible for the one-year “Follow-up Self-Report Survey” based on being discharged from their most recent

treatment episode by May 31, 2021, and lost to follow-up was defined as nonresponse to the one-year follow-up survey among eligible flight attendants. ^b974 of the 1,172 treatment episodes in the database were eligible for the one-year follow-up survey based on discharge dates by May 31, 2021. ^cDue to missing data, Ns ranged from 483 to 648. ^dDue to missing data, Ns ranged from 208 to 268. ^eMATs = medication-assisted treatments. ^fThis question was asked only of those who were offered MATs, so the percentages are based on a denominator N of 463 and 199 for the two groups, respectively. ^gIQR = interquartile range. ^hShows responses of “agree” or “strongly agree.”

SOURCE: Data from FADAP database and results from committee’s “Call for Perspectives.”

of age or biological sex. Flight attendants who were lost to follow-up were slightly, but non-significantly, less likely to have completed the “Initial Self-Report Survey” than flight attendants who were not lost to follow-up (86.6% vs. 91.1%). The analysis of loss to follow-up was limited, in that information was not available about severity of the presenting problem, race, ethnicity, and precise location, all factors shown to affect follow-up in other research (Cleland et al., 2004).

In terms of treatment episodes, 974 episodes were eligible for the one-year “Follow-up Self-Report Survey,” and 70.3 percent (N = 685) of these were lost to follow-up (see Table 5-1). Treatment episodes that were lost to follow-up generally did not differ from treatment episodes that were not lost to follow-up in terms of treatment-provider-reported treatment characteristics, offers of medication-assisted treatment, length of treatment, or treatment completion, and flight attendant reports from the “Initial Self-report Survey” (see Table 5-1).

Characteristics of Flight Attendants in FADAP¹

Table 5-2 shows the characteristics of the 1,196 unique flight attendants who took part in FADAP. Flight attendants were, on average, 44.4 years old (standard deviation = 11.6 years). Most flight attendants were female (57.5%, N = 686), which is unsurprising because the majority of all flight attendants are female. Nearly half of flight attendants were employed by one of two airlines—Airline 1 (25.0%, N = 299) and Airline 2 (20.9%, N = 250)—which are some of the largest airlines and, therefore, would be expected to show higher flight attendant representation in FADAP. However, some airlines were underrepresented in FADAP given their size² or for other reasons that are not clear. For example, one of the largest airlines,

___________________

¹ After a prepublication version of the report was provided to the FAA, this section was edited to anonymize the airline names.

² For more information, see Table A1 from FADAP Database Report in the Resources section of the report website.

TABLE 5-2 Characteristics of 1,196 Unique Flight Attendants Who Took Part in FADAP

NOTE: ^aThree flight attendants were missing data for biological sex.

SOURCE: Data from FADAP database and results from committee’s “Call for Perspectives.”

Airline 10, with 92,459 full-time employees in October 2022, was the most underrepresented in FADAP. Only 10 Airline 10 flight attendants took part in FADAP, which represents 0.01 percent of all Airline 10’s full-time employees and 0.84 percent of the 1,196 flight attendants in the FADAP database.³

FADAP Treatment Episode Details

Table 5-3 shows treatment episode details as reported by treatment providers in the “Primary Treatment Summary” survey. Notably, the unit of analysis is now treatment episodes (N = 1,172), and not flight attendants, because some flight attendants (11.6% [N = 139 of 1,196 unique flight attendants]) had more than one treatment episode recorded in the database. Only treatment episodes for which a treatment provider reported both

___________________

³ Information was pulled directly from an airline’s website, but the FAA requested to anonymous all airlines.

TABLE 5-3 Treatment-Provider-Reported Treatment Details for the 1,172 FADAP Treatment Episodes

NOTES: ^aDue to missing data, Ns ranged from 884 to 1,172 treatment episodes. ^bSpecific treatment issues are not mutually exclusive because treatment episodes could involve more than one specific issue. Therefore, percentages do not sum to 100. ^cIQR = interquartile range.

SOURCE: Data from FADAP database and results from committee’s “Call for Perspectives.”

admission and discharge dates are reported (N = 1,172). Due to missing data on some questions, analytic Ns ranged from 884 to 1,172 treatment episodes.

Approximately half of treatment episodes involved treatment for alcohol and mental health problems (51.7%, N = 574 of 1,110 treatment episodes). Almost one-third of treatment episodes (29.6%, N = 329 of 1,110 treatment episodes) involved treatment for a drug-use problem. Mental health problems were a common focus of treatment, particularly depression and anxiety, which were recognized by treatment providers as concerns in 44.3 percent (N = 519) and 35.8 percent (N = 419) of treatment episodes, respectively.

Medication-assisted treatment was offered in 92.6 percent of treatment episodes (N = 819 of 884 treatment episodes).⁴ Treatment was completed in 92.0 percent of treatment episodes (N = 994 of 1,081 treatment episodes). The median length of treatment was 31 days (interquartile range = 29, 43).

FADAP Effectiveness and Satisfaction

Table 5-4 shows flight attendants’ self-reported “recovery” status and satisfaction with FADAP at one year after treatment completion. The unit of analysis is treatment episodes. Treatment episodes with a discharge date by October 31, 2021, were considered eligible for the one-year follow-up survey (N = 990), given that the database was received in November 2022. Due to high rates of nonresponse by flight attendants to the one-year survey, analytic Ns ranged from 211 to 243 (i.e., 21.3–24.5% of 990 eligible episodes). Given the small N and limited ability to compare nonresponders and responders, there is a high degree of uncertainty about selection or attrition bias.

Flight attendants reported currently being in “recovery” at one-year post-treatment for 97.6 percent (N = 206) of treatment episodes and not in recovery for 2.4 percent (N = 5) of treatment episodes. Notably, this finding is based on only 211 (21.3%) of 990 eligible treatment episodes due to high rates of nonresponse to the one-year survey. Of the treatment episodes for which flight attendants reported being in recovery (N = 206), 95.6 percent (N = 197) were in continuous recovery for at least 30 days. Of the 197 treatment episodes that were in continuous recovery for at least 30 days, 68.3 percent (N = 125) of the episodes were characterized by a report of returning to work at one-year post-treatment.

___________________

⁴ Note that ~25 percent of the 1,172 treatment episodes (N = 288) were missing data on whether medication-assisted treatment was offered. FADAP revised the question about medication-assisted treatment in 2020, and when data from 2020 onward were used, medication-assisted treatment was offered in 83.9 percent of treatment episodes (256 of 305).

TABLE 5-4 Flight Attendant-Reported Recovery and Return-to-Work Status and Satisfaction with FADAP at One-Year Post-Treatment for 990 Treatment Episodes Eligible for the One-Year Post-Treatment Survey

NOTES: ^aThe analyses are limited to 990 treatment episodes that were eligible for the one-year follow-up survey based on discharge date by October 31, 2021. Due to high rates of nonresponse among flight attendants with eligible treatment episodes, analytic Ns ranged from 211 to 243, with the exception of the analysis of return to work. ^bFlight attendants were asked: “Would you describe yourself as currently being in recovery from both alcohol and drugs of abuse?” ^cThe question about return to work was asked only of the flight attendants who reported being in continuous recovery for at least 30 days. ^dShows responses of “agree” or “strongly agree” on a five-point Likert scale.

SOURCE: Data from FADAP database and results from committee’s “Call for Perspectives.”

In terms of satisfaction with FADAP at one-year after treatment discharge, satisfaction ratings were high and ranged from 81.1 to 88.5 percent reporting “agree” or “strongly agree” to statements such as “I would use FADAP again” on a five-point scale ranging from “strongly disagree” to “strongly agree.” However, the analytic N for satisfaction with FADAP ratings was 243 of 990 eligible episodes (24.5% of eligible episodes).

Approximately 16 percent of all unique treatment episodes (N = 182) were relapse episodes, defined as treatment episodes that occurred at least 28 days after the discharge date of a previous treatment episode.

Agreement of Findings: A Comparison Between FADAP Annual Reports and the Independent Analysis

Findings from the 2022 FADAP annual report, which reports on all data collected since 2014, were summarized and compared with findings from the independent analysis (Nordberg, 2023). There was general agreement on the findings. However, whereas the FADAP annual report concludes from the data that the program is effective, the committee’s interpretation of the data is more cautious. Low response rates to the one-year follow-up survey could bias outcomes such as recovery, return to work, and satisfaction with FADAP in a healthier/more positive direction.

Findings from the “Call for Perspectives” and Qualitative Interviews

The National Academies of Sciences, Engineering, and Medicine Committee developed a tool for gathering lived experiences (“Call for Perspectives”) and commissioned a qualitative study to understand the perspectives of flight attendants and pilots with regard to FADAP and HIMS, respectively. The National Academies of Sciences, Engineering, and Medicine contracted with Jennifer P. Wisdom (PhD, MPH, ABPP) of Wisdom Consulting to conduct the investigation. Here we summarize the results of the investigation. Wisdom’s full report, including copies of the “Call for Perspectives” and interview guide, can be found in the “Resources” section on the report’s website.

The “Call for Perspectives” is an online tool the committee used to gather lived experiences, posted on the National Academies’ website and disseminated via social media and mailing lists. Additionally, the committee enlisted the help of FADAP, HIMS, the Air Line Pilots Association, International, and other contacts within aviation to directly invite flight attendants and pilots to complete the “Call for Perspectives.” The committee recognizes this could have introduced selection bias into the results, but to ensure anonymity of the respondents, the link was not tracked. The “Call for Perspectives” asked about substance use and its relationship to airline industry culture, treatment initiation and recovery, and perspectives and recommendations on FADAP and HIMS. There were 1,188 respondents. Most “Call for Perspectives” respondents were flight attendants (99%, N = 1,173). Therefore, the results summarized here represent the views of flight attendants and not pilots.

The majority of respondents were women (69%, N = 812), and 27 percent were men (N = 322). (The remaining preferred not to answer [3%] or selected “non-binary/other.”) Most respondents had been employed in the industry for more than 10 years (63%, N = 746). A substantial percentage of respondents had no familiarity with the FADAP (33%, N = 395).

Approximately seven percent (N = 80) of respondents reported having previously been, or were currently, enrolled in FADAP. More than half of respondents stated that they would use FADAP if they thought they had a substance use disorder (52%, N = 616), and most respondents said they would recommend the program to a friend (91%, N = 84).

Purposive sampling was used to select a subset of “Call for Perspectives” respondents who indicated their willingness to complete qualitative interviews (N = 265 respondents). The sampling strategy sought to ensure diversity in profession (i.e., flight attendant versus pilot), sex, employment duration, and experience with the FADAP. Initial interview requests were sent to 40 purposely selected individuals. Given low response rates and time limitations, ultimately all 265 individuals who agreed in the “Call for Perspectives” to be contacted were invited to interview. Thirty-six interviews were conducted, and 35 of 36 interviewees were flight attendants. Therefore, the results from qualitative interviews reflect the views of flight attendants and not pilots.

Most interviewees were women (67%, N = 24), and most had more than 10 years of experience in the industry (75%, N = 27). Approximately 72 percent (N = 26) of interviewees had personal experience with substance use problems (80% alcohol; 16% psychostimulants; 8% sedatives; 4% GHB [gamma hydroxybutyrate, a depressant drug used as an intoxicant]; and 4% marijuana [percentages do not sum to 100% because of use of multiple substances]), and the remainder (N = 10) were not personally in recovery and not reporting substance misuse but had knowledge of a colleague or relative with a substance use disorder. Of 26 interviewees who had experience with substance misuse, all were in recovery; one had taken part in FADAP, six obtained treatment before FADAP existed, six sought treatment through the company or union Employee Assistance Program program, and the others sought treatment through other means (N = 10) or were unsure of how they got into treatment (N = 2). Three of 10 not in recovery were serving in a union role; 15 of the 26 who indicated they were in recovery were serving on their airline’s FADAP committee or in a union stewardship role. Several of the 11 individuals in recovery (out of the total of 26) who were not serving on these committees indicated a desire to serve. One interviewee entered treatment due to a positive U.S. Department of Transportation (DOT) test, whereas the others reported entering treatment voluntarily. Treatment was typically inpatient residential treatment followed by intensive outpatient treatment or mutual support groups. Several interviewees discussed hardships associated with returning to work and the need for greater treatment support on returning to work, especially to cope with triggers such as serving alcohol to passengers.

Interviewees identified aspects of working in the airline industry that increase risk for substance misuse, such as long hours, poor sleep, loneliness,

a culture of heavy drinking, and access to alcohol on the airplane and at airports. Interviewees revealed that some flight attendants do not recognize that they have a drinking problem, and, although crewmates are often aware of the problem, crewmates are hesitant to report it. Random testing by the DOT and the rule about no alcohol consumption eight hours prior to flying were reported to be ineffective substance-use deterrents. Interviewees mentioned the need for education and self-assessments about risky drinking to facilitate early intervention and treatment seeking.

Respondents to the “Call for Perspectives” and to the qualitative interviews clearly articulated the need for FADAP. However, they mentioned barriers to using FADAP, including a lack of awareness of FADAP, concerns about confidentiality, substance use stigma, fear of loss of employment (some airlines have a “zero tolerance” policy), and the financial costs of treatment. Suggested ways to improve FADAP included increasing promotion of FADAP to increase awareness of the program, dispelling misinformation to promote voluntary enrollment, enacting more rigorous substance testing requirements, and strengthening aftercare programs.

FINDINGS FOR HIMS

Findings from the HIMS Database

No independent analysis of the HIMS database was undertaken, because the committee’s repeated requests for access to the HIMS database were denied. HIMS’ claims about program effectiveness are summarized below, but because the committee lacked access to the relevant data, the claims could not be validated. To the committee’s knowledge, the claims have not been subjected to other third-party validation.

HIMS’ Self-reported Findings

HIMS asserts that the program is effective. Two sources for these claims are the HIMS executive summary and the HIMS 2021 Advanced Topics Seminar. The executive summary, which is dated December 2013 and is available on the HIMS website,⁵ states,

___________________

⁵ For more information, see https://himsprogram.com/wp-content/uploads/2021/04/ExecSummary.pdf

While the timeline of HIMS’ recovery measurements is unknown, this reported recovery rate is relative to the U.S. general population overall lifetime recovery rate of 75 percent (Jones et al., 2020).

A more recent version of the executive summary is not available on the website (checked March 10, 2023), but the following key points are still claimed:⁶

1. Between 1975 and 2022, more than 12,000 pilots had been helped, treated, and returned to their cockpits.
2. Eight hundred pilots recovering from an alcohol problem had achieved an 85 percent long-term abstinence rate.
3. A cost-benefit analysis showed a $9 return for every $1 spent on treatment.
4. The program enhanced flight safety.

In addition to the HIMS executive summary, claims about HIMS effectiveness were made in the HIMS 2021 Advanced Topics Seminar. The 2021 Advanced Topics Seminar data are shown in Tables 5-5, 5-6, and 5-7 and are reproduced exactly.

The seminar presented results from the HIMS database comprising first-class cases (N = 1,510 pilots, N = 1,291 Special Issuance Authorization) for whom an Airman Medical Examiner completed a datasheet from April 2011 to August 2021. Notably, the 1,510 pilots represent only 12.6 percent of the total 12,000 pilots HIMS claims have been returned to the cockpit from 1975–2021.

Table 5-5 shows the age ranges of the pilots and relapse rates by age. Most pilots in the report were between 40 and 59 years of age (70%). The overall relapse rate was 14 percent, and relapse rates were highest among the age 40–59 group (~16%) and generally lowest among the youngest age group (ages 20–29, 6.4%). It is difficult to evaluate what the low relapse rate means as it is unclear how relapse was defined or ascertained.

Table 5-6 shows pilots’ substance of choice and relapse rates by substance. Most pilots’ substance of choice was alcohol (92.5%), and the relapse rate for pilots whose substance of choice was alcohol was 13.7 percent. By comparison, few pilots’ substance of choice was opioids (2%), but the relapse rate for pilots whose substance of choice was opioids was 40 percent.

Table 5-7 shows how pilots entered HIMS. Most pilots entered HIMS through self-referral (28.4%) or through an off-duty driving under the influence (DUI) citation (24.1%), and a relatively small fraction of pilots (8.4%) entered HIMS through a positive drug test given by the DOT. It should be

___________________

⁶ Ibid.

TABLE 5-5 HIMS Pilots’ Ages and Relapse Rates by Age

NOTE: HIMS = Human Intervention Motivational Study. NR = not reported. ^aTotal N is shown in the HIMS presentation as 1,261, but actual N (based on summing pilots in each age category) is 1,264. Both Ns are under the number of pilots claimed to be in the database (N = 1,510 pilots, N = 1,291 Special Issuance Authorization).

SOURCE: Committee generated from the 2021 HIMS Advanced Topics Seminar (September 13–14, 2021, Denver, CO).

TABLE 5-6 HIMS Pilots’ Substance of Choice and Relapse Rates by Substance of Choice

NOTE: HIMS = Human Intervention Motivational Study. NR = not reported. Non-spec = non-specific. Semi-syn = semi-synthetic.

SOURCE: Committee generated from the 2021 HIMS Advanced Topics Seminar (September 13–14, 2021, Denver, CO).

noted that although “self-referral” sometimes entails a pilot independently recognizing that they need treatment without being under any particular workplace pressure to enter treatment at that time, in other cases pilots are advised to refer themselves to HIMS after they reported an incident (e.g., a DUI) that placed their medical certificate at risk.

TABLE 5-7 HIMS Pilot Program Entry

NOTE: HIMS = Human Intervention Motivational Study. DOT = Department of Transportation. TSA = Transportation Security Administration. ^aThese percentages are slightly off if the total N is assumed to be 1,692 pilots (the sum of pilots across entry mechanisms), which is greater than the sum total of pilots reported to be included in the database supported by the seminar. However, it is unclear whether pilots were allowed to be included in more than one entry mechanism category. ^bDOT Positive Drug Test.

SOURCE: Committee generated from the 2021 HIMS Advanced Topics Seminar (September 13–14, 2021, Denver, CO).

Findings from the “Call for Perspectives” and Qualitative Interviews

Almost no data were collected from pilots as part of the National Academies’ “Call for Perspectives” tool to gather experiences (N = 15 pilots vs. 1,157 flight attendants) or as part of the National Academies’ commissioned study involving qualitative interviews (N = 1 pilot vs. N = 35 flight attendants). Unlike FADAP, the committee never received indications that HIMS and its administering organization, Air Line Pilots Association, International, ever distributed the link or sought pilot participation. The data collected from a very few pilots are not reported separately here to preserve anonymity.

This page intentionally left blank.