Abstract
Background:
The performance of quantitative pupillary light reflex (qPLR) and the neurological pupil index (NPi) was used to predict neurological outcomes in cardiac arrest (CA) patients.
Methods:
Eligible studies on the ability of the qPLR and NPi to predict neurological outcomes in CA patients were searched from the PubMed and China National Knowledge Infrastructure databases until July 2023. The pooled odds ratio (OR) and its 95% confidence interval (95% CI), area under the curve, sensitivity analysis, and publication bias were analyzed via Stata 14.0 software.
Results:
Twelve studies involving 1530 CA patients (968 in the qPLR study and 1025 in the NPi study) satisfied the inclusion criteria. qPLR (OR: 24.50, 95% CI: 13.08–45.86%, P < .001) and NPi (OR: 15.55, 95% CI: 7.92–30.55%, P < .001) were significantly associated with neurological outcome. The pooled area under the curves of the qPLR and NPi for predicting neurological outcome were 0.89 (95% CI: 0.86–0.92%) and 0.66 (95% CI: 0.62–0.70%), respectively. The pooled results were robust according to the sensitivity analysis. The funnel plots for qPLR (P = .94) and NPi (P = .10) did not reveal any publication bias.
Conclusion:
Compared with the NPi, the qPLR has greater diagnostic accuracy in predicting favorable neurologic outcomes.
Keywords: cardiac arrest, neurological outcome, neurological pupil index, quantitative pupillary light reflex
1. Introduction
Cardiac arrest (CA) is a medical emergency condition that requires immediate management to save lives and is an important cause of death worldwide, especially in middle-income or developing countries. Despite developments in resuscitation techniques and an increased rate of successful resuscitation, most patients with CA are discharged from the hospital with permanent brain damage.[1,2] Accurate predictions may aid health care professionals in providing appropriate care to patients and valid information to their relatives.
Several studies have associated changes in the pupillary response with neurological outcomes in CA patients, which suggests a neurological relationship with the quantitative pupillary light reflex (qPLR) and the neurological pupil index (NPi).[3–14] However, these studies yielded inconsistent results, leading to controversies regarding the clinical significance of the pupillary response in predicting neurological outcomes. This systematic review and meta-analysis aimed to provide a direction for the development and update of the use of the qPLR and NPi in the assessment of neurological outcomes.
2. Materials and methods
The systematic review and meta-analysis were performed according to the preferred reporting items for systematic reviews and meta-analysis statements. All involved studies passed ethical approval, and informed consent was obtained.
2.1. Literature search
Electronic databases, including the PubMed and China National Knowledge Infrastructure databases, were systematically searched from the date of their inception until July 2023. The keywords used for the search included “cardiac arrest” with “pupillary light reflex,” “neurological pupil index,” or “quantitative pupillometry.” The reference lists of the identified literature were further used to manually search for potentially relevant literature.
2.2. Inclusion and exclusion criteria
Two reviewers searched the abovementioned literature databases separately and screened them according to the inclusion and exclusion criteria. The inclusion criteria were as follows: (1) study type: observational studies (cohort, case–control, and cross-sectional studies), (2) patients: study subjects were adults with CA, and (3) index test: qPLR or NPi for predicting neurological outcome. The exclusion criteria were as follows: (1) animal experiments, (2) studies containing duplicate data or repeated publications, and (3) studies with incorrect data or no relevant results.
2.3. Data extraction
Two investigators independently extracted the data of each eligible study via a standardized form including the first author, publication year, study location, study design, sample size, intervention method, and outcome measures. Investigators could contact the original author by email or phone to obtain missing but crucial study-related information when needed. Disagreements were resolved through discussions with a third investigator.
Neurological outcomes were assessed via the Glasgow–Pittsburgh Cerebral Performance Categories (CPCs). Patients were classified as having a favorable neurologic outcome (CPC 1: full recovery, CPC 2: moderate disability) or an unfavorable neurologic outcome (CPC 3: severe disability, CPC 4: persistent vegetative state, and CPC 5: death).[3]
2.4. Quality assessment
The risk of bias assessment was independently assessed by 2 researchers via the Quality Assessment of Diagnostic Accuracy Studies-2 scale, which mainly consists of 4 parts: case selection, trial evaluation, the gold standard, case flow, and progress. Disagreements were checked by a third reviewer.
2.5. Statistical analysis
The meta-analysis of all the data was performed via Stata 14.0 software (StataCorp, USA). The I2 statistic was used to assess heterogeneity. I2 > 50% or P < .05 was considered indicative of significant heterogeneity. The random effects model was used in cases of considerable heterogeneity, and the fixed effects model was used for low heterogeneity. Forest plots showing odds ratios (ORs) with corresponding 95% confidence intervals (CIs) for each study and the pooled estimate of overall effects were generated. Subgroup analysis was applied to assess the effects of the specific study characteristics on the diagnostic accuracy of qPLR or NPi. Sensitivity analyses were also performed according to the methodological quality parameters. Publication bias was assessed among studies via Deek funnel plot asymmetry tests. P < .05 was considered statistically significant.
3. Results
3.1. Literature search
The study selection procedure used in the meta-analysis is shown in Figure 1. Across the 3 databases, 76 results were retrieved, of which 14 were duplicates. After the titles and abstracts were scanned, the full texts were read, and the reference lists of the articles were manually searched. Twelve studies, which included a total of 1530 CA patients, were ultimately included to evaluate the associations between the qPLR and NPi with neurological outcomes.[3–14]
Figure 1.
PRISMA flowchart of the literature search.
3.2. Study characteristics
The characteristics of the 12 studies[3–14] included in this review are listed in Table 1. The sample size ranged from 40 to 456 and included 1530 CA patients with an average age of 61 years. The median percentage of unfavorable neurological outcomes of the 12 studies[3–14] selected was 58%, with a range of 26% to 74%. Eight studies[6–12,14] investigated qPLR, and 7 studies[3–8,13] explored NPi. Neurological outcomes were assessed via the CPC score in all studies.
Table 1.
Characteristics of included studies.
| Author (year) | Age | Male (%) | Study region | Study design | Sample size | OHCA (%) | Poor neurological outcome (%) | Measure method | Measure time | Neurological outcome |
|---|---|---|---|---|---|---|---|---|---|---|
| Ellouze (2020) | 62 | NA | France | Prospective | 40 | 18 | 70 | qPLR | NA | CPC at 3 months |
| Heimburger (2016) | 61 | 82 | France | Prospective | 82 | 79 | 67 | qPLR | Day 2 | CPC at 3 months |
| Macchini (2022) | 66 | 66 | Belgium | Retrospective | 102 | 53 | 68 | qPLR + NPi | Day 1 | CPC at 3 months |
| Menozzi (2022) | 57 | 73 | Europe | Prospective | 66 | NA | 65 | NPi | Admission | CPC at 3 months |
| Nyholm (2023) | 62 | 81 | Denmark | Retrospective | 135 | 100 | 39 | qPLR + NPi | Day 1 | CPC at 3 months |
| Oddo (2018) | 62 | 78 | Europe | Prospective | 456 | NA | 59 | qPLR + NPi | Day 1 | CPC at 3 months |
| Paramanathan (2021) | 58 | 88 | Europe | Retrospective | 65 | 100 | 26 | NPi | Day 3 | CPC at 6 months |
| Peluso (2022) | 62 | 75 | Europe | Prospective | 329 | NA | 61 | NPi | Admission | CPC at 3 months |
| Riker (2020) | 57 | 65 | USA | Prospective | 52 | 80 | 69 | qPLR + NPi | Day 1 | CPC at discharge |
| Solari (2017) | 62 | 69 | Switzerland | Prospective | 103 | 100 | 51 | qPLR | Day 2 | CPC at 1 years |
| Suys (2014) | 61 | 68 | Switzerland | Prospective | 50 | 100 | 54 | qPLR | Day 1 | CPC at 3 months |
| Tamura (2018) | 63 | 72 | Japan | Prospective | 50 | 100 | 74 | qPLR | Admission | CPC at 3 months |
CPC = cerebral performance categories, NA = not available, NPi = neurological pupil index, OHCA = out-of-hospital cardiac arrest, qPLR = pupillary light reflex.
3.3. Quality assessment
The methodological quality assessment for the studies included in this meta-analysis is shown in Figure S1, Supplemental Digital Content, http://links.lww.com/MD/O299. The source of risk in the index test could be attributed to the threshold that was not prespecified and nonconsecutive patients. However, all included studies met the review questions; thus, the applicability judgments were of low concern.
3.4. qPLR for neurological outcomes
Eight studies involving 968 patients provided data regarding the association between the qPLR and neurological status among CA patients.[6–12,14] The heterogeneity of the qPLR between the included studies (I2 = 0.0%, P = .639) was not significant. The qPLR (OR: 24.50, 95% CI: 13.08–45.86%, P < .001) was significantly associated with neurological outcome (Fig. 2). The pooled area under the curve (AUC) of the qPLR for predicting neurological outcome was 0.89 (95% CI: 0.86–0.92%), with 96% specificity and 57% sensitivity (Fig. 3A).
Figure 2.
Forest plot of the ability of the quantitative pupillary light reflex to predict neurological outcome.
Figure 3.
Summary ROC curve of the quantitative pupillary light reflex (A) and neurological pupil index (B) for predicting neurological outcomes. AUC = area under the ROC curve, ROC curve = receiver operating characteristic curve, SENS = sensitivity, SPEC = specificity.
3.5. NPi for neurological outcome
Seven studies involving 1025 patients provided data regarding the association between NPi and neurological symptoms among CA patients.[3–8,13] The heterogeneity of the NPi between the included studies (I2 = 0.0%, P = .463) was not significant. The NPi (OR: 15.55, 95% CI: 7.92–30.55%, P < .001) was significantly associated with neurological outcome (Fig. 4). The pooled AUC of the NPi for predicting neurological outcome was 0.66 (95% CI: 0.62–0.70%), with 99% specificity and 24% sensitivity (Fig. 3B).
Figure 4.
Forest plot of the ability of the neurological pupil index to predict neurological outcomes.
3.6. Subgroup analysis
The effects of specific study characteristics on the diagnostic accuracy of the qPLR for predicting neurological outcomes were assessed in this study.[6–12,14] In particular, subgroup analysis was conducted based on average age (<62 or ≥62), male percentage (<75% or ≥75%), average sample size (<120 or ≥120), cutoff (prespecified or not prespecified), study design (prospective or retrospective), and measurement time (before day 1 or after day 1). These variables may affect the robustness of the meta-analysis results. Subgroup analysis further showed that these study characteristics did not substantially affect the effect estimates (Table 2).
Table 2.
Subgroup analysis of qPLR for predicting neurological outcome.
| Characteristics | OR (95% CI) | P-value | I2 (P-value) |
|---|---|---|---|
| Age | |||
| <62 | 10.61 (3.88–29.02) | P < .001 | 0% (P = .886) |
| ≥62 | 34.65 (15.20–78.98) | P < .001 | 0% (P = .855) |
| Male percentage | |||
| <75% | 19.73 (8.38–46.46) | P < .001 | 5.3% (P = .348) |
| ≥75% | 33.70 (12.39–91.64) | P < .001 | 0% (P = .508) |
| Sample size | |||
| <120 | 22.41 (11.22–44.76) | P < .001 | 7.0% (P = .494) |
| ≥120 | 28.83 (8.10–102.62) | P < .001 | 0% (P = .517) |
| Cutoff | |||
| Prespecifed | 35.21 (13.42–92.39) | P < .001 | 0% (P = .589) |
| Not prespecified | 15.24 (6.75–34.39) | P < .001 | 0% (P = .599) |
| Study design | |||
| Prospective | 22.65 (11.89–43.16) | P < .001 | 0% (P = .619) |
| Retrospective | 66.43 (3.87–1139.28) | P = .004 | / |
| Measure time | |||
| Before day 1 | 25.01 (10.61–58.96) | P < .001 | 0 % (P = .882) |
| After day 1 | 23.05 (8.70–61.09) | P < .001 | 73.5% (P = .052) |
CI = confidence interval, OR = odds ratio, qPLR = quantitative pupillary light reflex.
The effects of specific study characteristics on the diagnostic accuracy of the NPi for predicting neurological outcomes were also assessed.[3–8,13] In particular, subgroup analysis was conducted based on average age (<62 or ≥60), male percentage (<75% or ≥75%), average sample size (≥170 or <170), cutoff (prespecified or not prespecified), study design (prospective or retrospective), and measurement time (before day 1 or after day 1). These variables may affect the robustness of the meta-analysis results. Subgroup analysis also showed that these study characteristics did not substantially affect the effect estimates (Table 3).
Table 3.
Subgroup analysis of NPi for predicting neurological outcome.
| Characteristics | OR (95% CI) | P-value | I2 (P-value) |
|---|---|---|---|
| Age | |||
| <62 | 9.62 (4.55–20.36) | P < .001 | 0% (P = .627) |
| ≥62 | 43.29 (8.44–222.02) | P < .001 | 0% (P = .404) |
| Male percentage | |||
| <75% | 6.16 (2.25–16.81) | P < .001 | 0% (P = .889) |
| ≥75% | 25.12 (9.77–64.58) | P < .001 | 0% (P = .529) |
| Sample size | |||
| <170 | 8.13 (3.40–19.47) | P < .001 | 0% (P = .829) |
| ≥170 | 28.45 (9.01–77.63) | P < .001 | 55.4% (P = .134) |
| Cutoff | |||
| Prespecifed | 10.41 (4.48–24.16) | P < .001 | 0% (P = .879) |
| Not prespecified | 27.31 (8.69–85.85) | P < .001 | 58.8% (P = .099) |
| Study design | |||
| Prospective | 15.72 (7.48–33.09) | P < .001 | 45.8% (P = .138) |
| Retrospective | 14.70 (2.99–72.39) | P = .001 | 0% (P = .910) |
| Measure time | |||
| Before day 1 | 15.87 (7.71–32.66) | P < .001 | 28.7% (P = .230) |
| After day 1 | 13.23 (1.98–88.58) | P = .008 | 0% (P = .674) |
CI = confidence interval, NPi = neurological pupil index, OR = odds ratio.
3.7. Sensitivity analysis and publication bias
Sensitivity analysis confirmed the stability of the pooled OR, and no individual study affected the overall result (Fig. 5). No publication bias was found in the funnel plot for qPLR (P = .94, Fig. 6A) or NPi (P = .10, Fig. 6B).
Figure 5.
Sensitivity analysis of the quantitative pupillary light reflex (A) and neurological pupil index (B) for predicting neurological outcomes.
Figure 6.
Publication bias of the included studies analyzed via Deek method: (A) quantitative pupillary light reflex and (B) neurological pupil index.
4. Discussion
The interaction between the PLR and neurological outcome is complicated and has raised concerns in recent years. This systematic review and meta-analysis included results from 10 studies with a total of 1530 patients to assess the prognostic performance of the qPLR and NPi for the prediction of unfavorable neurological outcomes after CA. The pooled results suggest that the qPLR has greater diagnostic accuracy in predicting favorable versus unfavorable neurologic outcomes than does the NPi. In addition, the qPLR and NPi have near-perfect pooled specificities of 100% and relatively low performances, with pooled sensitivities of 57% and 24%, respectively.
The PLR is the intrinsic mechanism of pupil constriction in response to changing light levels and is modulated by cognitive brain function and interaction with the superior colliculus, pretectal olivary nucleus, and locus coeruleus.[15–17] Pupillary dilatation commonly results from increased intracranial pressure that causes oculomotor nerve entrapment due to uncal herniation across the tentorial incisura, which signals ongoing or impending mechanical compression of the brainstem.[18] In addition to oculomotor nerve compression, blood flow imaging has shown that pupillary changes are highly correlated with brainstem oxygenation and perfusion/ischemia.[19] Research has shown decreased brainstem function following CA and resuscitation, which might explain why the PLR can reflect neurological function.
Wang et al[20] conducted a meta-analysis and demonstrated that the qPLR was related to neurological outcome and had an AUC of 0.83, which is consistent with the results of 0.89 in our study. Compared with the study of Wang et al,[20] our study performed a sensitivity analysis, which resulted in more persuasive and stronger conclusions. Kim et al[21] evaluated the usefulness of the qPLR as a prognostic tool for neurological outcomes in post-CA patients treated with targeted temperature management, and the results did not reveal the effectiveness of the qPLR in post-CA patients treated with TTM (AUC = 0.75). The qPLR might have been influenced by hypothermia because the sphincter pupillae muscle produces a small decrease in the light reflex amplitude.[22]
Notably, the pooled sensitivity of the qPLR and NPi for predicting neurological outcome was relatively low. Four studies[3,5,6,8,10] considered optimizing the Youden index for a specificity of 90% to 100%, which resulted in a sensitivity of 9% to 61%. Clinicians consider the withdrawal of life-sustaining therapy, and the tests used for prognostication must have extremely high specificities to avoid withdrawal of life-sustaining therapy in patients with plausible chances of survival. Another possible explanation for the low sensitivity of qPLR and NPi on admission could be related to gradually slowing pupillary recovery after reperfusion. The early neuronal dysfunction that occurs during anoxic injury in some cerebral areas sensitive to hypoxia, such as in the brainstem or retina,[23,24] could result in fixed mydriasis. The return of spontaneous circulation (ROSC) and brain reperfusion could result in some neuronal recovery, with pupils becoming reflective of light stimulation.[4] Although the sensitivity of ultrasonic prediction is relatively low, the non-invasiveness, accessibility, immediate feedback, and irrespection of temperature and sedation offered by point-of-care ultrasound render it indispensable for assessing neurological outcomes after CA.[11] In the future, the integration of a multimodal prognostication approach might increase the overall predictive accuracy.
The neurological examination of the PLR is a key component of the prognostic assessment of CA patients, but the optimal measurement time has not yet been established. The neurological outcome deteriorates rapidly during the first few days after CA, but some comatose patients may recover despite injury.[25] The prognostic performance of the qPLR might have decreased from days 1 to 3, except for the NPi, which improved the AUC across days.[3,5,6] Tamura et al[12] reported that the immediate qPLR after ROSC best predicted favorable neurological outcomes (AUC = 0.84) and decreased to 0.61 to 0.68 from days 1 to 3. Thus, qPLR is feasible as early as within an hour after ROSC.
This study comprised a comprehensive search and extensive analysis and the use of current guidelines for reviewing and assessing the bias of prediction studies. However, several factors still limit the statistical power of the data. First, the included studies did not use the same protocol, and end points varied across studies. Second, the lack of unified measurement criteria, cutoff values, and measurement times for the qPLR and NPi resulted in the use of various judgment criteria in the eligible studies. Although subgroup analysis showed that the measurement criteria, cutoff value, and measurement time did not substantially change the effect estimates, the generalizability of the qPLR and NPi needs to be considered with caution. Therefore, a common protocol for measuring these parameters needs to be proposed. Third, some of the included studies were single-center and retrospective. The absence of randomization and blinding might have also increased the risk of bias. In the future, prospective multicenter and randomized studies are needed to assess the diagnostic performance of the qPLR and NPi. Fourth, the number of included studies was relatively small, and insufficient information regarding follow-up and comorbidities limited more specific segment analyses in this review. Finally, as most of the original included studies provided only 3-month outcomes, we determined the value of the qPLR and NPi for the prediction of long-term outcomes (e.g., at 6 months or 1 year).
The results of our systematic review and meta-analysis showed that the qPLR has greater diagnostic accuracy in predicting favorable versus unfavorable neurologic outcomes than the NPi does. Given the limited number of studies included in this review, more prospective, high-quality cohorts and the integration of a multimodal prognostication approach are needed to improve neurologic outcomes after CA.
Acknowledgments
The author thank science assistants for the data collection and assessment of bias.
Author contributions
Conceptualization: Chang Sheng Feng.
Methodology: Chang Sheng Feng.
Writing – original draft: Chang Sheng Feng.
Writing – review & editing: Chang Sheng Feng.
Supplementary Material
Abbreviations:
- AUC
- area under the curve
- CA
- cardiac arrest
- CI
- confidence interval
- CPC
- cerebral performance categories
- NPi
- neurological pupil index
- OR
- odds ratio
- qPLR
- pupillary light reflex
- ROSC
- return of spontaneous circulation
The authors have no funding and conflicts of interest to disclose.
All data generated or analyzed during this study are included in this published article [and its supplementary information files].
Supplemental Digital Content is available for this article.
How to cite this article: Feng CS. Performance of the quantitative pupillary light reflex and neurological pupil index for predicting neurological outcomes in cardiac arrest patients: A systematic review and meta-analysis. Medicine 2025;104:4(e41314).
References
- [1].Dumas F, Bougouin W, Perier MC, et al. Long-term follow-up of cardiac arrest survivors: protocol of the DESAC (Devenir des survivants d’Arrets Cardiaques) study, a French multicentric prospective cohort. Resusc Plus. 2023;16:100460. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [2].Srinivasan V, Hall J, Wahlster S, Johnson NJ, Branch K. Associations between clinical characteristics of cardiac arrest and early CT head findings of hypoxic ischaemic brain injury following out-of-hospital cardiac arrest. Resuscitation. 2023;190:109858. [DOI] [PubMed] [Google Scholar]
- [3].Macchini E, Bertelli A, Bogossian EG, et al. Pain pupillary index to prognosticate unfavorable outcome in comatose cardiac arrest patients. Resuscitation. 2022;176:125–31. [DOI] [PubMed] [Google Scholar]
- [4].Menozzi M, Oddo M, Peluso L, et al. Early neurological pupil index assessment to predict outcome in cardiac arrest patients undergoing extracorporeal membrane oxygenation. ASAIO J. 2022;68:e118–20. [DOI] [PubMed] [Google Scholar]
- [5].Paramanathan S, Grejs AM, Søreide E, et al. Quantitative pupillometry in comatose out-of-hospital cardiac arrest patients: a post-hoc analysis of the TTH48 trial. Acta Anaesthesiol Scand. 2022;66:880–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [6].Oddo M, Sandroni C, Citerio G, et al. Quantitative versus standard pupillary light reflex for early prognostication in comatose cardiac arrest patients: an international prospective multicenter double-blinded study. Intensive Care Med. 2018;44:2102–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [7].Riker RR, Sawyer ME, Fischman VG, et al. Neurological pupil index and pupillary light reflex by pupillometry predict outcome early after cardiac arrest. Neurocrit Care. 2020;32:152–61. [DOI] [PubMed] [Google Scholar]
- [8].Nyholm B, Obling LER, Hassager C, et al. Specific thresholds of quantitative pupillometry parameters predict unfavorable outcome in comatose survivors early after cardiac arrest. Resusc Plus. 2023;14:100399. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [9].Heimburger D, Durand M, Gaide-Chevronnay L, et al. Quantitative pupillometry and transcranial Doppler measurements in patients treated with hypothermia after cardiac arrest. Resuscitation. 2016;103:88–93. [DOI] [PubMed] [Google Scholar]
- [10].Solari D, Rossetti AO, Carteron L, et al. Early prediction of coma recovery after cardiac arrest with blinded pupillometry. Ann Neurol. 2017;81:804–10. [DOI] [PubMed] [Google Scholar]
- [11].Suys T, Bouzat P, Marques-Vidal P, et al. Automated quantitative pupillometry for the prognostication of coma after cardiac arrest. Neurocrit Care. 2014;21:300–8. [DOI] [PubMed] [Google Scholar]
- [12].Tamura T, Namiki J, Sugawara Y, et al. Quantitative assessment of pupillary light reflex for early prediction of outcomes after out-of-hospital cardiac arrest: a multicentre prospective observational study. Resuscitation. 2018;131:108–13. [DOI] [PubMed] [Google Scholar]
- [13].Peluso L, Oddo M, Sandroni C, Citerio G, Taccone FS. Early neurological pupil index to predict outcome after cardiac arrest. Intensive Care Med. 2022;48:496–7. [DOI] [PubMed] [Google Scholar]
- [14].Ellouze O, Constandache T, Berthoud V, Voizeux P, Bouhemad B, Guinot P-G. Automated infrared pupillometry for neurological prognostication after extracorporeal cardiopulmonary resuscitation. Intensive Care Med. 2020;46:549–51. [DOI] [PubMed] [Google Scholar]
- [15].Hall CA, Chilcott RP. Eyeing up the future of the pupillary light reflex in neurodiagnostics. Diagnostics (Basel). 2018;8:19. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [16].Ciuffreda KJ, Joshi NR, Truong JQ. Understanding the effects of mild traumatic brain injury on the pupillary light reflex. Concussion. 2017;2:CNC36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [17].Miyamoto K, Murakami I. Pupillary light reflex to light inside the natural blind spot. Sci Rep. 2015;5:11862. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [18].Chen JW, Gombart ZJ, Rogers S, Gardiner SK, Cecil S, Bullock RM. Pupillary reactivity as an early indicator of increased intracranial pressure: The introduction of the Neurological Pupil index. Surg Neurol Int. 2011;2:82. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [19].Marshman LA. Brain stem blood flow, pupillary response, and outcome in patients with severe head injuries. Neurosurgery. 2002;51:848–9; author reply 849. [PubMed] [Google Scholar]
- [20].Wang CH, Wu CY, Liu CC, et al. Neuroprognostic accuracy of quantitative versus standard pupillary light reflex for adult postcardiac arrest patients: a systematic review and meta-analysis. Crit Care Med. 2021;49:1790–9. [DOI] [PubMed] [Google Scholar]
- [21].Kim JG, Shin H, Lim TH, et al. Efficacy of quantitative pupillary light reflex for predicting neurological outcomes in patients treated with targeted temperature management after cardiac arrest: a systematic review and meta-analysis. Medicina (Kaunas). 2022;58:804. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [22].Larson MD, O’Donnell BR, Merrifield BF. Ocular hypothermia depresses the human pupillary light reflex. Invest Ophthalmol Vis Sci. 1991;32:3285–7. [PubMed] [Google Scholar]
- [23].Miroz JP, Ben-Hamouda N, Bernini A, et al. Neurological pupil index for early prognostication after venoarterial extracorporeal membrane oxygenation. Chest. 2020;157:1167–74. [DOI] [PubMed] [Google Scholar]
- [24].Hofmeijer J, Mulder AT, Farinha AC, van Putten MJAM, le Feber J. Mild hypoxia affects synaptic connectivity in cultured neuronal networks. Brain Res. 2014;1557:180–9. [DOI] [PubMed] [Google Scholar]
- [25].Clark DL, Penner M, Orellana-Jordan IM, Colbourne F. Comparison of 12, 24 and 48 h of systemic hypothermia on outcome after permanent focal ischemia in rat. Exp Neurol. 2008;212:386–92. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.