Automated, quantitative pupillometry in the critically ill: a systematic review of the literature: online supplementary data
Opic, Petra (2021), Automated, quantitative pupillometry in the critically ill: a systematic review of the literature: online supplementary data, Dryad, Dataset, https://doi.org/10.5061/dryad.1c59zw3v7
Objective: A systematic literature search has been performed to identify potential confounders for outcome prediction using pupillary light reflex in adult critically ill patients, as measured by handheld automated pupillometry devices.
Methods: Three digital databases (PubMed, EmBase, Cochrane) were systematically searched. Articles published between 1990-2019 in adult patients, using monocular automated handheld devices were considered. Studies were classified according to the OCEBM classification (level 1 represents the highest, level 5 the lowest level of evidence). Case reports, original research and systematic reviews were included and cross-referenced.
Results: Using 202 search terms, 58 eligible articles reporting on the use of handheld pupillometry in the critically ill could be identified, considering 3246 patients. The highest level of evidence came from 10 randomized trials and 19 prospective observational studies. The level of evidence was mostly 2 to 3 and highest with studies regarding the potential confounding effects of pain, the use of opioids and increased intracranial pressure. Additional potential confounders found are SSRIs, α2-adregenic receptor agonists and NMDA antagonists.
Conclusions: The pupillary light reflex is susceptible to factors resulting from underlying comorbidities and effects of treatment regimes. Scenarios frequently encountered in critical care, such as pain, the use of opioids, and proof of increased intracranial pressure have potential confounding effects on outcome and pupillary reflexes. When treatment is guided by pupillary metrics, such confounders put patients at risk of over- or undertreatment. Future research should validate and identify additional confounders, as our review suggests that even more unexplored confounders may exist.