Pilots using selective serotonin reuptake inhibitors compared to other fatally injured pilots

Accident Analysis and Prevention 107 (2017) 86–91

Contents lists available at ScienceDirect

Accident Analysis and Prevention journal homepage: www.elsevier.com/locate/aap

Pilots using selective serotonin reuptake inhibitors compared to other fatally injured pilots

MARK

⁎

Paul Rogersa, , Christy Hilemana, Guillermo Salazara, Kacey Cliburna, Lawrence Paskoffa, William Hathawaya, Kevin Gildeaa, Victor Hugo Tejera Villalazb a b

Federal Aviation Administration, Civil Aerospace Medical Institute, United States National University of Columbia, Colombia

A R T I C L E I N F O

A B S T R A C T

Keywords: SSRI Aviation safety Alcohol Antihistamine Antidepressant Depression

Selective Serotonin Reuptake Inhibitors (SSRI) were a disqualifying medication for U.S. civil pilots before April 5, 2010. After this date, a Federal Aviation Administration policy was created that allowed airmen, on select SSRIs, a pathway to hold a valid medical certificate. The purpose of this study was to provide a detailed look at SSRIs in the U.S. pilot population since the inception of this new policy. We examined the toxicology results from fatally injured airmen in addition to outcomes concerning pilots who are participating in the program. This study examined data from the Civil Aerospace Medical Institute's Bioaeronautical Sciences Research Laboratory in conjunction with the Medical Analysis Tracking Registry and the Document Imaging and Workflow System. A count-based regression model quantified the relationships between positive SSRI findings with additional factors of interest. These factors included pilot rating, ethanol, and first generation antihistamines. There were 1484 fatally injured airmen over the six year study period, of which 44-tested positive for an SSRI. First-generation antihistamines were statistically associated with positive findings of SSRIs.

1. Introduction The Centers for Disease Control and Prevention (CDC) reported that antidepressant use in the United States increased nearly 400% between periods spanning 1988–1994 and 2005–2008 (Pratt et al., 2011). Overall, they were the third most common prescription medication in use from 2005 to 2008 by Americans of all ages, and the most frequently used drug in the 18–44 age group. It was reported from 2005 to 2008 that 11% of all Americans over the age of 12 were taking an antidepressant. In the 2012 CDC National Ambulatory Medical Survey, out of the 20 categories to which 3583 physicians responded, antidepressants and anxiolytics/sedatives/hypnotics were the third and fourth most prescribed medications (CDC-NCHS, 2012). The increasing prevalence of antidepressant use in the U.S. general population raises a number of questions concerning the U.S. pilot population. These questions are of interest to policy makers, regulators, and the flying public. Before April 2010, the use of antidepressants was considered disqualifying for pilots seeking a Federal Aviation Administration (FAA) airman medical certificate. The disqualification was due to the medication and/or the underlying medical condition being treated. The FAA

conducted an extensive review of the experiences of Transport Canada and the Civil Aviation Safety Authority in Australia assessing the risks and benefits of antidepressant usage in civil aviation (Ross et al., 2007; Federal Aviation Administration (US), 2010; Jones and Ireland, 2004). Starting April 5, 2010, the FAA permitted an Authorization for Special Issuance of a Medical Certificate (Authorization) for airmen on selected Selective Serotonin Reuptake Inhibitors (SSRI), a group of antidepressants better tolerated by patients than older tricyclic (TCA) drugs (Federal Aviation Administration, 2017). The selected SSRIs were fluoxetine (Prozac®), sertraline (Zoloft®), citalopram (Celexa®) and escitalopram (Lexapro®) (Berry, 2010). Allowing a Special Issuance for selective SSRIs signified a shift in long-standing FAA policy regarding antidepressants and deserves closer study. In order to qualify for this Authorization an airman may only have had one of the following diagnoses: 1. major depressive disorder that was mild to moderate and could have been single or recurrent; 2. dysthymic disorder; 3. adjustment disorder; or 4. any non-depression related condition for which an SSRI would be needed. Furthermore, the condition needed to have been stable for a minimum of six continuous months prior to the application for an Authorization and on an established dose of medication. Finally, the applicant must never have had

⁎

Corresponding author. E-mail addresses: [email protected] (P. Rogers), [email protected] (C. Hileman), [email protected] (G. Salazar), [email protected] (K. Cliburn), lawrence.paskoff@faa.gov (L. Paskoff), [email protected] (W. Hathaway), [email protected] (K. Gildea), [email protected] (V.H. Tejera Villalaz). http://dx.doi.org/10.1016/j.aap.2017.07.023 Received 25 September 2016; Received in revised form 20 June 2017; Accepted 23 July 2017 0001-4575/ Published by Elsevier Ltd.

Accident Analysis and Prevention 107 (2017) 86–91

P. Rogers et al.

modeling of rare events. Therefore, a count-based Poisson regression model was employed to explore the relationship of SSRIs with factors related to the aviators’ instrument rating, antihistamines, and ethanol. The Poisson distribution can be defined in terms of a single parameter (λ) as:

any of the following: psychosis, suicidal ideation, electro-convulsive therapy, treatment with multiple SSRIs concurrently, or multi-agent drug protocol use (i.e., prior use of other psychiatric medications in conjunction with SSRIs). The airman would engage a specially designated Human Intervention Motivation Study (HIMS) Aviation Medical Examiner (AME) to assist in submitting all required information, and if all these criteria were met the Federal Air Surgeon would consider the airman for an Authorization for Special Issuance of a medical certificate for continued use of the SSRI (Berry, 2010; Federal Aviation Administration, 2017). SSRIs were frequently found in conjunction with other medications during toxicological testing of biological samples from aircraft crash victims. A previous study covering the period 1990–2001 of 61 fatally injured airmen who tested positive for an SSRI, found that 39 of the 61 had other drugs present (Akin and Chaturvedi, 2003). Diphenhydramine, a first-generation antihistamine, was the most common drug found in fatally injured pilots (Canfield et al., 2011). Ethanol was also a widely used drug, frequently discovered in the samples taken from fatally injured pilots; one study reported finding ethanol in 7% of this group (Chaturvedi et al., 2016). Alcohol and first-generation antihistamines are considered impairing substances, which may work synergistically with the side effects of SSRIs. Both alcohol and antihistamines are central nervous system depressants with adverse effects on neuronal activity resulting in sedation and cognitive impairment (Dry et al., 2012; Fox et al., 2014; Woehrling et al., 2015). The use of diphenhydramine together with fluoxetine, sertraline, escitalopram, or citalopram may increase side effects such as dizziness, drowsiness, and difficulty concentrating. Also, there was evidence that alcohol and SSRIs, in combination, lead to unexpected behaviors (Menkes and Herxheimer, 2014). The aim of this study was to provide a statistical and descriptive analysis to identify factors and trends associated with the use of SSRIs within the fatally injured civil pilot population. Further, this study was conducted to determine if it is possible to quantify and describe the relationship of SSRI use with other crash factors within this population. These other factors included pilot instrument rating, ethanol, and firstgeneration antihistamines. This information will be of use in the formulation of recommendations concerning SSRIs within the aviation community.

f (k ; λ ) =

e−λλk , k = 0, 1, 2,...... k!

(1.0)

Where λ typically represents the rate, in terms of event occurrence, and k is the number of these events. The event (k) is the discovery of a SSRI in the toxicology of a fatally injured airman. Eq. (2.0) describes the dependent variable in terms of the log counts of fatally injured accident airmen: ln [Count (SSRI Airmen)] = β0 + β1 *(Instrument Rating ) + β2 *(Ethanol)+ β3 *(Antihistamine ) + β4 *(Antihistamine *Ethanol) + ln (offset )

Another predictor in the Poisson model, the offset or exposure, does not have a regression coefficient to be estimated. The offset represents the denominator, or total number of airmen, in a particular category or covariate pattern. We needed to include this offset to calculate the results in terms of Incident Rate Ratios (IRR) within the regression model. One of the fundamental assumptions in Poisson regression is that the mean and variance are equal: λ = μ was a necessary condition for producing valid standard errors for the regression coefficients. Slight departures from this assumption can be compensated for with the use of a dispersion parameter used to scale the standard errors (Hilbe, 2014). In this study, the dispersion parameter was set equal to Pearson’s ChiSquare statistic divided by its degrees of freedom to adjust the model’s standard errors. The model produced results in terms of IRRs in units of Person-Years for individual rates. The statistically insignificant variables were removed in a backward elimination process assessing interaction terms before their main effects. All analyses were performed in Statistical Analysis Software (SAS) version 9.4. The level of significance for all tests was set at α ≤ 0.05. 2.1. Variable categorization and classification 2.1.1. SSRI The count of SSRI cases was the outcome variable of interest. The total number of fatalities was represented by the offset in the model for each covariate pattern.

2. Methods In this project, the population of interest was fatally injured aviators; this group was selected as toxicology and autopsy information were readily available and required to confirm the presence or absence of specific drugs or medications. The FAA performs toxicology and collects autopsy information on all pilots fatally injured in general aviation crashes. In addition, this research relied upon the Medical Analysis Tracking (MANTRA) Registry system maintained by the Autopsy Program Team located at the Civil Aerospace Medical Institute (CAMI). MANTRA is a nexus for varied sources of information on fatal aerospace incidents. These sources included the National Transportation Safety Board (NTSB), Document Imaging Workflow System, Airmen Registry, toxicology results, and the autopsy reports submitted from medicolegal death investigations. The study timeframe spanned from October 1, 2008 through September 30, 2014 corresponding to the 2009–2014 government fiscal years. The SSRIs examined in this research included the following: SSRIs allowable under the FAA SSRI policy (citalopram, escitalopram, fluoxetine, and sertraline) and other non-approved SSRIs (fluvoxamine, paroxetine, vilazodone, indalpine, and zimelidine). We queried MANTRA for findings involving any of the listed SSRIs. Once matched, we pulled the entire case for a more detailed examination. In order to quantify the relationships of the data in the various systems we used regression techniques. Typically, analyzing the occurrence of adverse events in the aviation environment involves the

2.1.2. Instrument The variable Instrument indicated if the airmen held an instrument rating. An instrument rating is the additional training required to fly under instrument flight rules and was intended to be a surrogate variable for both proficiency and as an experience measure. It was created as a binary variable, with a one indicating that the airman held an instrument rating and a zero otherwise. The counts of this variable were then summed for each category. 2.1.3. Ethanol Ethanol is an impairing drug frequently found in the fatally injured airmen population. However, ethanol may also be produced postmortem by bacteria; therefore, it was extremely important to determine the source of the ethanol. The presence of certain substances, such as npropanol, n-butanol and serotonin metabolites, might suggest postmortem production, but this must be considered with care. In a study by Chaturvedi et al. spanning years 1989–2013, 85 of the 1169 cases tested positive for ethanol (Chaturvedi et al., 2016). Of these 85 cases that tested positive for ethanol, six (7.1%) of these findings were deemed to have been due to post-mortem production of ethanol. This finding provided a rough estimate of the rate at which samples were affected by the post-mortem production of ethanol. If the proportion of 87

Accident Analysis and Prevention 107 (2017) 86–91

P. Rogers et al.

rotorcraft), 26 were single engine land (SEL) and three were multiengine land (MEL). Of the LSA pilots, two had valid medical certificates, three had no certificates and four had certificates that were expired (Table 2). The two glider pilots had expired medical certificates. Of the SEL pilots, 22 had valid medical certificates, one had a certificate that had been denied and three had certificates that had expired. Lastly, all three MEL pilots had valid medical certificates. The NTSB classified two of the 44 (4.6%) incidents as caused by suicide (NTSB Numbers: CEN11FA130, ERA13FA330). This percentage of suicides was greater than the 0.29% reported by Lewis for all aviation incidents in a study spanning years 2003–2012 (Lewis et al., 2015). Thirty-four (2.29%) of the overall sample size were female; only one of the 44 fatally injured pilots was female. The median age of all pilots included in the study was 56 years. The median age of the subgroup of pilots who tested positive for an SSRI was 61 and ranged from 22 to 82 years. In a National Health and Nutrition Examination Survey, concerning prescription drug use among U.S. adults spanning years 1999–2012, 15% of the 2352 participants in age groups of 40–64 years used five or more prescription medications at a time (Kantor et al., 2015). In comparison, seven (16%) airmen in this study used five or more prescriptions. These findings suggest that airmen killed in crashes are a comparable subset of the general population with similar medical conditions. As of January 1, 2016 a total of 293 airmen received an Authorization to fly while taking an SSRI. None of these airmen were among the 44 fatal accidents involving SSRI positive airmen.

fatally injured airmen undergoing the post-mortem production of ethanol was the same in the two groups of aviators who tested positive and negative for a SSRI then the effects of this post-mortem production of ethanol will cancel out in the IRRs produced by the model. That is, the post-mortem production of ethanol will not bias the estimate of the regression coefficient. The variable Ethanol was a dichotomous variable with zero and one indicating the absence and presence of the drug, respectively. The counts were summed for each covariate pattern in the model. 2.1.4. Antihistamine Antihistamine was a variable summed over the various categories of the model indicating the presence of first-generation antihistamines. First-generation antihistamines included chlorpheniramine, clemastine, diphenhydramine, hydroxyzine, tripolidine, and brompheniramine. 2.2. Power analysis and model methodology The statistical power of the Poisson regression model described in Eq. 2.0 was dependent on a number of factors to include the significance level and effect size. Initially, we desired to have statistical power at 80% for an effect size involving at least a 10% difference in the ratio of incidence rates. Our power calculations were based on the work by Signorini (Signorini, 1991). At a significance level(α)of 0.05 and using an incidence rate ratio of 1.08 (8% difference) as an effect size, with a sample size of 1484 cases, the estimated statistical power was 84%. In terms of statistical power and effect size, the model was considered viable.

3.1. Distribution of frequencies The factors listed in Eq. 2.0 were categorized by positive and negative SSRI test results in order to generate the distribution of frequencies reported in Table 3.

2.3. Crash examination The details of each fatal incident in which a SSRI positive airman was involved were examined in depth. Information concerning the medical certification status, operational Federal Aviation Regulations, accident location, relevant medications, aircraft type, and NTSB probable cause were taken into account and summarized.

3.2. Statistical analysis – quantifying risk The model’s Deviance statistic indicated the specified Poisson regression fit the data reasonably well (χ2 = 2.8629; DF = 3; pvalue = 0.4133). The interaction term in Eq. 2.0 between Ethanol and Antihistamine was found to be statistically insignificant; the covariates Instrument and Ethanol were also found to be statistically insignificant. The variable Antihistamine (Wald χ2 = 5.15, p = 0.02) was the only statistically significant factor in the model. The IRR for the Antihistamine covariate was IRR = 2.54. This result indicated that fatally injured aviators who tested positive for an SSRI were twice as likely to have a first generation antihistamine in their system over pilots who tested negative for an SSRI (Table 4).

3. Results The data extracted from MANTRA yielded 1484 fatally injured airmen over the period covered in this study. Of these 1484 fatalities, 44 airmen had positive test results for at least one SSRI yielding a prevalence of 2.9% of all fatally injured airmen. The 44 cases that tested positive for a SSRI were distributed by drug as shown in Table 1. CAMI forensic toxicology testing did not distinguish between the stereoisomers of citalopram and escitalopram and positive results were reported as citalopram. Of the airmen listed in Table 1, none had been issued an Authorization. Counts for controlled and performance-impairing pharmaceuticals not listed in this table were zero. Within the group of airmen with positive toxicology results for an SSRI, 40 of the incidents occurred under Title 14, Code of Federal Regulations (FAR) Part 91 rules (general operating and flight rules); one each under FAR Part 135 (commuter/on demand) and Part 137 (agricultural) procedures; two occurred under FAR Part 103 (ultralight) guidelines. Nine of the aircraft were classified as light sport (LSA), two were gliders, four were rotorcraft (one was a Part 103 ultralight

3.3. Ethanol prevalence Although statistically insignificant, the overall prevalence of Ethanol was high and merits further examination. Table 3 indicated that 30% (446/1484) of all fatally injured airmen tested positive for ethanol. The prevalence of ethanol in this population seems much higher than the estimate of 7% reported by Chaturvedi et al. (Section 2.1.3) (Chaturvedi et al., 2016). The difference between this study’s results and that of Chaturvedi et al. can be accounted for by the threshold in determining a positive ethanol finding. In the study by Chaturvedi et al., a concentration greater than or equal to 40 mg/dl was classified as a positive finding. They chose 40 mg/dl, since under the FAA regulation, this is a blood alcohol concentration at which no person may operate or attempt to operate an aircraft (e-CFR, 2017). In this study, any result for ethanol including those less than 40 mg/dl was considered a positive finding. When we filtered our ethanol cases at 40 mg/dl the overall prevalence was reduced to 6.8% (102/1484). This value was very close to the 7%

Table 1 Distribution of SSRI cases by drug. SSRI

Cases

Citalopram/Escitalopram Fluoxetine Sertraline Paroxetine

20 6 15 3

88

Accident Analysis and Prevention 107 (2017) 86–91

P. Rogers et al.

Table 2 A summary of medical certificate status by aircraft class.a NTSB Number

Age

Gender

FAR

Aircraft Type

Number of Medications

SSRI

Medical Certificate

CEN12FA271 N/A CEN12FA638 CEN12LA203 ERA12FA491 CEN13LA244 WPR13FA269 CEN14LA372 CEN14LA454 ERA13LA076 N/A CEN12LA460 CEN14LA419 ERA11LA056 ERA14FA115 ERA14FA461 CEN09PA348 ERA11FA219 ERA14FA123 WPR11LA272 CEN11LA569 WPR11FA188 CEN12FA170 CEN09FA462 CEN09FA562 CEN09LA466 CEN11FA075 CEN11FA130 ERA10FA414 ERA11FA095 ERA11LA022 CEN11FA358 ERA11FA232 CEN12FA611 ERA12FA458 ERA12FA526 ERA12FA566 ERA12FA583 ERA13FA330 ERA13FA348 CEN14LA202 ERA14FA328 WPR14FA226 CEN12LA209

82 64 68 74 79 62 77 67 67 82 66 72 45 54 49 61 35 50 58 65 67 73 63 59 47 46 55 50 61 56 56 57 51 67 79 50 61 48 22 39 63 58 72 57

M M M M M M M M M M M M M M M M M F M M M M M M M M M M M M M M M M M M M M M M M M M M

91 103 91 91 91 91 91 91 91 91 103 91 91 91 91 91 91 135 91 137 91 91 91 91 91 91 91 91 91 91 91 91 91 91 91 91 91 91 91 91 91 91 91 91

SEL Glider SEL LSA LSA LSA SEL Glider LSA SEL Rotor LSA LSA LSA Rotor Rotor Rotor MEL LSA SEL SEL SEL SEL SEL SEL LSA SEL SEL SEL SEL SEL SEL MEL SEL SEL SEL SEL SEL SEL MEL SEL SEL SEL SEL

4 5 6 8 3 7 3 4 4 7 2 2 2 4 4 2 1 2 1 2 2 2 4 5 1 2 4 1 2 3 1 2 4 4 5 1 2 2 2 1 2 2 2 2

SERTRALINE SERTRALINE CITALOPRAM CITALOPRAM CITALOPRAM SERTRALINE FLUOXETINE CITALOPRAM CITALOPRAM FLUOXETINE CITALOPRAM PAROXETINE PAROXETINE SERTRALINE CITALOPRAM SERTRALINE FLUOXETINE SERTRALINE CITALOPRAM SERTRALINE SERTRALINE SERTRALINE PAROXETINE SERTRALINE CITALOPRAM CITALOPRAM CITALOPRAM CITALOPRAM CITALOPRAM SERTRALINE SERTRALINE FLUOXETINE CITALOPRAM CITALOPRAM SERTRALINE FLUOXETINE CITALOPRAM CITALOPRAM CITALOPRAM SERTRALINE CITALOPRAM CITALOPRAM SERTRALINE FLUOXETINE

Denied Medical Expired Certificate Expired Certificate Expired Certificate Expired Certificate Expired Certificate Expired Certificate Expired Certificate Expired Certificate Expired Certificate No Certificate No Certificate No Certificate No Certificate Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid Valid

a

The classes single- and multi-engine land are abbreviated SEL and MEL, respectively.

3.4. Additional crash details

Table 3 Distribution of frequencies between SSRI test results and the model covariates.

Each incident is a unique event made up of a number of different elements. We provided a brief description of the 44 fatal crashes involving positive SSRI findings in Table 2 in an effort to provide a brief overview of the investigation details. The NTSB numbers for the 42 cases investigated are provided to aid the reader in finding the details of these incidents. Two crashes, out of the total of 44, were classified as ultralights and not investigated by the NTSB. Of the 44 fatalities that tested positive for an SSRI, 37 (84%) had multiple medications in their system. Airmen utilized SSRIs in conjunction with other medications, making it highly probable that they had other co-morbidities in addition to depression or anxiety. The presence of an SSRI was typically a marker other substances were present in that airman and some of these may have played a role in the crash. In Table 2, we totalled the number of medications discovered in each case. At the time of this writing, the NTSB had issued final reports on 37 of the 42 cases it investigated. The NTSB noted medical issues in 28 out of the 37 (76%) cases positive for an SSRI. However, the lack of a determination by the NTSB concerning the causal effects of a medical condition and/or medication in a crash does not preclude that it may have been a significant factor. The NTSB is not able to

SSRI Variable Instrument Rated Not Rated Ethanol Present Absent Antihistamine Present Absent

Positive

Negative

Totals

% Positive

21 23 44

760 680 1440

781 703 1484

2.69 3.27 2.96

16 28 44

430 1010 1440

446 1038 1484

3.59 2.70 2.96

7 37 44

93 1347 1440

100 1384 1484

7.00 2.67 2.96

reported by Chaturvedi et al. The toxicology laboratory reported values as low as 10 mg/dl therefore, 23% of this study’s positive findings were between 10 and 40 mg/dl.

89

Accident Analysis and Prevention 107 (2017) 86–91

P. Rogers et al.

Table 4 Poisson regression results for antihistamine covariate. Variable

Estimate

Std. Error

Wald 95% Confidence Limits

Wald Chi-Square

P-Value

IRR

Antihistamine

0.9339

0.4404

0.0708

4.50

0.0339

2.5444

1.7971

involved in a fatal crash. Although perceived cumbersome by some airmen, the process appears to ensure that applicants were receiving constructive oversight enhancing flight safety. The regression model identified first-generation antihistamines as being associated with the counts of fatal crashes in which there was a positive test for an SSRI. This association can be explained by the frequent use of multiple medications in this subgroup of aviators. Airmen who failed to divulge their SSRI use on their medical application were likely to be using multiple medications to treat several co-morbidities that could impact flight safety. The limitations of this study include the inability to examine each non-SSRI case individually to determine the number of medications in use by these individuals. To examine the number of medications, as well as the medication type and dosage, would require a case-by-case review by an experienced toxicologist and physician. This was not feasible in this study. In conclusion, airmen who declared their use of SSRIs and participated in the process of obtaining an Authorization from the FAA pose minimal risk to the National Airspace System. In addition, the discovery of an undeclared SSRI in a fatally injured airmen was an infrequent occurrence. There was no direct evidence that SSRIs, by themselves, endangered flight safety.

determine the human performance impact of many medical conditions and/or medications. 4. Discussion In general, fatal aviation crashes were an infrequent occurrence. That is, less than 1% of all U.S. general aviation pilots experience an incident and, of this number, less than 21% of them involved a fatality (NTSB, 2014). Out of the 1484 fatalities in this study, only 44 tested positive for an SSRI. This amounted to less than 3% of all fatally injured pilots during this timeframe. From these numbers, we concluded that, in aviation as a whole, a fatal aviation accident with a positive SSRI toxicology result was a rare event. The FAA developed a process in April 2010 for airmen utilizing selected SSRI antidepressants to be considered for an Authorization. In this study population, none of the 44 fatally injured airmen reported the use of those medications as required by the FAA (Federal Aviation Administration (US), 2010). The reason for this is unknown but it may be that some pilots considered this path too cumbersome. The process required applicants to provide a written statement describing their use of the antidepressant. They were also to provide treatment records for antidepressant use, a report from their treating/prescribing physician on their diagnoses and past medication use, an evaluation by a psychiatrist, and a neuropsychological evaluation with the results of a CogScreen – Aeromedical Edition test. If they had held a first- or second-class airman medical certificate and had flown for a commercial carrier they must provide a letter from company management (Chief Pilot or designee) attesting to their competence, crew interaction, and mood. Lastly, they were required to see and provide a detailed evaluation by a HIMS AME. The burden for providing these evaluations and reports must be borne by the applicant and may be a deterrent to considering the special issuance process. Many, in the cohort of fatally injured airmen in this study, failed to disclose one or more medications and psychiatric/psychological conditions. The FAA medical application specifically asked those questions concerning these medications and conditions in items 17a and 18 of the Application for an Airman Medical Certificate (FAA Form 8500-8). In item 20 of this form, the applicant is required to make a declaration of the completeness and truthfulness of his or her responses on the medical application. Even if the psychiatric/psychological condition and/or medication use started after the application, Title 14, Code of Federal Regulations Part 61.53(a) and 91.7(a) (3) still requires an individual to not exercise the privileges of their airman medical certificate. This lack of candid reporting or disregard for regulations may indicate a personality that chooses to ignore rules and regulations. Of the 44 airmen who tested positive for an SSRI, nine previously held a medical certificate and let it expire. The reasons were not clear but at least three would have required it for SEL operations. The other six may have let the certificate expire because they did not require it (glider, LSA) or may not have wanted to pursue the Authorization for their medical condition and/or medication use, an action that may have been a violation of existing FAA Regulations. There were 30 fatally injured airmen who held a valid medical certificate but chose to continue to fly even though they were on a medication and/or medical condition that would have been disqualifying. One SEL airman continued to fly even though his medical certificate had been denied and it was necessary for the type of aircraft he was flying. Through January 1, 2016, 293 airmen applied and obtained an Authorization for the use of an SSRI and none of these airmen were

References Akin, A., Chaturvedi, A.K., 2003. Prevalence of Selective Serotonin Reuptake Inhibitors in Pilot Fatalities of Civil Aviation Accidents, 1990–2001. U.S. Department of Transportation, Federal Aviation Administration, Office of Aerospace Medicine Report, Washington, DC (May Report No.: DOT/FAA/AM-03/7). Berry, M.A., 2010. New SSRI certification guidelines in place. Federal Air Surgeon's Medical Bulletin. 48 (2), 1–5. CDC-NCHS. Centers for Disease Control and Prevention (CDC), National Center for Health Statistics (NCHS), 2012. National center for health statistics (NCHS). National Ambulatory Medical Care Survey; State and National Summary Tables. ([Accessed on 5 May 2016] Available from: http://www.cdc.gov/nchs/data/ahcd/namcs_ summary/2012_namcs_web_tables.pdf). Canfield, D.V., Dubowski, K.M., Chaturvedi, A.K., Whinnery, J.E., 2011. Drugs and alcohol in civil aviation accident pilot fatalities from 2004–2008. U.S. Department of Transportation, Federal Aviation Administration, Office of Aerospace Medicine Report, Washington, DC (Sep. Report No.: DOT/FAA/AM-11/13). Chaturvedi, A.K., Craft, K.J., Hickerson, J.S., Rogers, P.B., Canfield, D.V., 2016. Ethanol and drugs found in civil aviation accident pilot fatalities: 1989–2013. Aerosp. Med. Hum. Perform. 87 (5), 470–476. Dry, M.J., Burns, N.R., Nettelbeck, T., Farquharson, A.L., White, J.M., 2012. Dose-related effects of alcohol on cognitive functioning. PLoS One 7 (11), 1–7. e-CFR. Electronic Code of Federal Regulations (e-CFR), 2017. Title 14: Aeronautics and Space, Part 91-General Operating and Flight Rules, Subpart A-General: §91.17 Alcohol or Drugs. ([Accessed on 5 May 2016.] Available from: http://www.ecfr.gov/ cgi-bin/text-idx?node=14:2.0.1.3.10). Federal Aviation Administration (US), 2010. Special issuance of airman medical certificates to applicants being treated with certain antidepressant medications. Fed. Regist. 75 (April (64)), 17047–17050 (5). Federal Aviation Administration, 2017. Guide for Aviation Medical Examiners. ([Accessed on 5 May 2016]. Available from: http://www.faa.gov/about/office_org/ headquarters_offices/avs/offices/aam/ame/guide/). Fox, C., Smith, T., Maidment, I., Chan, W., Bua, N., Myint, P., et al., 2014. Effect of medications with anti-cholinergic properties on cognitive function, delirium, physical function and mortality: a systematic review. Age Ageing 43, 604–615. Hilbe, J., 2014. Modeling Count Data. Cambridge Univ Press, New York (NY) (283 p.). Jones, D.R., Ireland, R.R., 2004. Aeromedical regulation of aviators using selective serotonin reuptake inhibitors for depressive disorders. Aviat. Space Environ. Med. 75 (5), 461–469. Kantor, E., Rehm, C., Haas, J., Chan, A., Giovannucci, E., 2015. Trends in prescription drug use among adults in the United States from 1999–2012. JAMA 314 (17), 1818–1831. Lewis, R., Forster, E., James, W., 2015. Aircraft-assisted pilot suicides in the United States, 2003-2012. J. Depress. Anxiety 4 (2). Menkes, D.B., Herxheimer, A., 2014. Interaction between antidepressants and alcohol: signal amplification by multiple case reports. Int. J. Risk Saf. Med. 26 (3), 163–170.

90

Accident Analysis and Prevention 107 (2017) 86–91

P. Rogers et al.

safety in civil aviation: a case-control study of 10 years of Australian data. Aviat. Space Environ. Med. 78 (8), 749–755. Signorini, D., 1991. Sample size for poisson regression. Biometrika 78 (2), 446–450. Woehrling, E., Parri, H., Tse, E., Hill, E., Maidment, I., Fox, G., et al., 2015. A predictive in vitro model of the impact of drugs with anticholinergic properties on human neuronal and astrocytic systems. PLoS One 10 (3), 1–18.

NTSB, 2014. Review of US Civil Aviation Accidents Calendar Year 2011. National Transportation Safety Board (NTSB), Washington, DC (Report No.: NTSB/ARA14-01.). Pratt, L.A., Brody, D.J., Gu, Q., 2011. Antidepressant Use in Persons Aged 12 and Over: United States, 2005–2008. NCHS Data Brief, No 76. National Center for Health Statistics, Hyattsville, MD. Ross, J., Griffiths, K., Dear, K., Emonson, D., Lambeth, L., 2007. Antidepressant use and

91