Abstract
Background: COVID-19 has infected millions of people worldwide, and acute COVID-19 survivors are experiencing long-term effects that affect their neurocognitive and mental health. Airline pilots are particularly vulnerable to even mild persistent COVID-19 impairments because of the demanding nature of their jobs. COVID-19 can cause attention deficits and executive dysfunction, impairing pilots' information processing, problem-solving, attention allocation, multi-tasking, and decision-making, particularly under stress or during long-haul flights that cross time zones.
Addressing the long-term neurocognitive impacts of COVID-19 on flight safety is challenging, especially if it is mild, undetected, or not reported. Despite widespread vaccination and the low risk of severe post-COVID-19 conditions among airline pilots, mild impairments critical for flight safety may still occur, and pilots may also be unaware of or unwilling to disclose their impairments. This doctoral thesis was designed to identify, evaluate, and characterise flight safety risk-mitigation strategies for addressing the long-term neurocognitive effects of COVID-19 on airline pilots. Objectively identifying neurocognitive impairments in pilots recovering from COVID-19 is essential as a way of selecting likely impaired subjects for further comprehensive neuropsychological evaluations.
Objective: The overarching objective of this PhD research programme was to identify and characterise objective screening tools suitable for detecting possible neurocognitive impairment in pilots following COVID-19 infections.
Methods: Literature synthesis on post-COVID-19 conditions and long-term effects, neurocognitive impairments and their screening tests for pilot performance, and critical pilot performance parameters were conducted through systematic searches and narrative reviews. Effective screening tests for various neurocognitive function domains were identified and evaluated for their predictive value regarding the demanding skills of airline pilots and their sensitivity in screening pilots. The results were sent to international subject matter experts for evaluation and validation, using a modified Delphi method, involving two rounds.
The psychometric properties of the identified screening tests in individuals with post-COVID-19 conditions and healthy controls, and their acceptability and feasibility in pilot populations, were then evaluated using a cross-sectional mixed-methods study. Quantitative data were collected via a survey questionnaire, and qualitative data were obtained through interviews. Statistical analyses, such as the Shapiro-Wilk Test, Mann-Whitney U test, Spearman’s rank correlation, multiple regression, and the receiver operating characteristic curve, were performed, using IBM SPSS version 29.0.
Results: The initial evidence review identified a range of neurocognitive and neuropsychiatric effects of COVID-19, including physical and mental fatigue, which can affect an individual's performance. Post-COVID-19 patients, including those with minor symptoms and those who were unaware of their impairments, experienced impairments in attention and executive function. From synthesising neuropsychological determinants and factors influencing pilot performance, essential neurocognitive skills, such as attention and executive function, are critical for safe flight performance in pilots.
From 13 neurocognitive screening tools that were identified for detecting COVID-19-related impairments across various cognitive domains, the digital versions of the Trail Making Test parts A and B and the Symbol Digit Coding version of the Symbol Digit Modalities Test were selected and validated as quick and suitable screening tools in airline settings. These tools have aviation norms, are objective, have good utility, are available in multiple versions, assess cognitive abilities relevant to pilot performance, and are economical for aviation medical examiners.
The area under the receiver operating characteristic curve for the Trail Making Test Part B was 0.85 [p<0.001; 95%CI=0.755-0.952], followed by the Symbol Digit Coding version of the Symbol Digit Modalities Test 0.81 [p<0.001; 95%CI=0.708-0.927], and the Trail Making Test Part A 0.76 [p=0.01; 95%CI=0.938-0.900]. The Trail Making Test Part B had a sensitivity of 71.9%, specificity of 91.3%, and overall accuracy of 80%. The Symbol Digit Coding version of the Symbol Digit Modalities Test had a sensitivity and specificity of 53.1% and 95.6%, respectively, with an overall accuracy of 70.9%. The Trail Making Test Part A showed a sensitivity of 71.9%, specificity of 78.3%, and overall accuracy of 74.5%. The Mann-Whitney U test indicated significant score differences between cases and controls in all tests, with a large effect size for Trail Making Test Part B [r=-0.59; Z=-4.43; p<0.001] and the Symbol Digit Coding version of the Symbol Digit Modalities Test [r=0.53; Z=3.98; p<0.001], and a medium effect size for Trail Making Test Part A [r=-0.45; Z=-3.37; p<0.001].
From the qualitative part of the study, five key themes emerged: the impacts of post-COVID-19 conditions on driving; job-related impacts; distinct features, trends, and long-term consequences of post-COVID-19 conditions; lack of diagnosis and management of post-COVID-19 conditions; and stakeholders’ lack of focus on post-COVID-19 conditions.
All tests were deemed acceptable and credible for use in routine aeromedical screening. Participant pilots viewed neurocognitive screening in routine aeromedical evaluations as a beneficial strategy for them and the aviation industry. They found the instructions for each screening test easy to understand and follow, encountering no challenges or technical difficulties while performing the tests, as the screening tests were neither burdensome nor time-consuming.
Conclusions: Post-acute COVID-19 impairments have been observed across various cognitive domains. Impairments in attention and executive function were widely reported and could be critical to flight safety if pilots are unaware of or are unwilling to disclose these impairments. Airline pilots found both the Trail Making Test parts A and B and the Symbol Digit Coding version of the Symbol Digit Modalities Test acceptable and feasible for aeromedical use.
The Trail Making Test Part B demonstrated greater sensitivity than the Symbol Digit Coding version of the Symbol Digit Modalities Test and showed higher specificity than the Trail Making Test Part A. Moreover, the Trail Making Test Part B exhibited the highest discriminative power and accuracy, and this is crucial to sufficiently detecting subtle post-COVID-19 neurocognitive impairments, but significant enough to pose flight safety risks. Furthermore, the Trail Making Test Part B is a quick, suitable, effective, and reasonably practicable test for assessing neurocognitive impairments, alongside other medical and operational evaluations.
This screening test, therefore, provides a way of selecting likely impaired airline pilots for further comprehensive neuropsychological evaluation. A large-scale study conducted at a national or international level is warranted, preferably to establish an aircrew-specific normative dataset and to examine additional psychometric properties of the selected screening tool.