Impact of ophthalmic draping on end-tidal CO₂, cerebral oximetry, and optic nerve sheath diameter in cataract surgery: an observational comparison by COVID-19 pneumonia history with randomized oxygen flow allocation

Abstract

Background

Ophthalmic draping during cataract surgery creates an isolated surgical field but may lead to carbon dioxide (CO₂) accumulation beneath the drape. Elevated CO₂ can cause hypercapnia, cerebral vasodilation, and increased intracranial pressure, potentially reflected by changes in cerebral oximetry and optic nerve sheath diameter (ONSD). Patients with a history of COVID-19 pneumonia may exhibit residual pulmonary dysfunction that influences these parameters during surgery. This study aimed to investigate the effects of ophthalmic draping on end-tidal CO₂, cerebral oximetry, and ONSD in patients undergoing cataract surgery and to determine whether these effects differ between patients with and without a history of COVID-19 pneumonia.

Methods

This prospective comparative study included 80 ASA I–III patients aged > 18 years with and without a history of COVID-19 pneumonia who underwent unilateral phacoemulsification under local anesthesia. Participants were divided into two main groups: those with (Group A, n = 40) and without (Group B, n = 40) a history of COVID-19 pneumonia. Within each predefined COVID-19 history stratum, patients were randomly allocated by block randomization to receive either 2 L/min (A-2, B-2) or 4 L/min (A-4, B-4) oxygen insufflation beneath the surgical drape. End-tidal CO₂ and hemodynamic parameters were continuously monitored, while cerebral oximetry parameters (rSO₂, ΔO₂Hb, ΔHHb, ΔcHb) were measured bilaterally every five minutes. ONSD was assessed ultrasonographically before and after surgery.

Results

ETCO₂ values showed no significant group-by-time differences. Postoperative ONSD values were significantly higher than preoperative values within both main groups and all oxygen-flow subgroups; however, the between-group comparisons were not statistically significant. Right frontal relative deoxyhemoglobin changes (ΔHHb-right) differed significantly between Groups A and B at the 10th minute (p = 0.003), while the other cerebral oximetry parameters demonstrated no significant group- or time-dependent differences. No patient developed hypoxia or neurological deterioration.

Conclusions

In patients undergoing short-duration unilateral phacoemulsification under local anesthesia with supplemental oxygen at 2–4 L/min, ophthalmic draping was not associated with clinically meaningful differences in ETCO₂ or ONSD, regardless of the history of COVID-19 pneumonia. Although selected cerebral oximetry parameters showed statistically significant differences between groups, all measured respiratory, hemodynamic, and cerebral physiological variables remained within physiological limits throughout the procedure. These findings should be interpreted within the context of the specific clinical settings studied.

Trial registration

ClinicalTrials.gov, NCT05571683. Retrospectively registered on 7 October 2022.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12871-026-04213-z.

Introduction

Cataracts, recognized as a major public health issue worldwide, are most commonly treated using the phacoemulsification technique [1–3]. During phacoemulsification, a sterile drape is routinely applied to isolate the eyelashes and periocular area, which may serve as potential sources of pathogens. Because these surgical drapes are impermeable to air, the accumulation of carbon dioxide (CO₂) beneath the drape can occur intraoperatively [4–7]. If the patient’s end-tidal CO₂ (ETCO₂) level increases during surgery, compensatory hyperventilation and tachycardia are expected [8]. Moreover, elevation of partial arterial CO₂ pressure (PaCO₂) may induce intracerebral vasodilatation, leading to increased intracranial pressure, which can be reflected as an enlargement of the optic nerve sheath diameter (ONSD), a dural extension of the intracranial compartment [9, 10].

Based on this physiopathological mechanism, we hypothesized that CO₂ accumulation beneath the ophthalmic drape during cataract surgery may cause increases in ETCO₂ values. The resulting intracerebral vasodilatation and elevation of intracranial pressure are expected to alter cerebral oxygenation parameters and increase ONSD measurements. ETCO₂ levels were defined as the primary outcome, while the secondary outcomes included cerebral oximetry parameters, optic nerve sheath diameter changes, and hemodynamic variables.

Previous studies have shown that ophthalmic draping may alter respiratory gas dynamics during cataract surgery performed under local anesthesia. Schlager reported that CO₂ accumulation beneath the drape can elevate ETCO₂ levels even in non-sedated patients without underlying respiratory dysfunction, while low-flow oxygen insufflation at 2 L/min prevents hypoxia but does not prevent hypercapnia [5]. Similarly, Schlager and Luger demonstrated that both hypoxia and CO₂ rebreathing can occur under ophthalmic drapes and that nasal oxygen administration reduces desaturation episodes but does not prevent CO₂ retention [6]. Risdall and Geraghty later confirmed that oxygen delivery via a nasal cannula at 2 L/min effectively prevents hypoxia during phacoemulsification and recommended routine oxygen supplementation for all patients undergoing ophthalmic surgery under local anesthesia [11].

However, despite these observations, the potential effects of drape-related CO₂ accumulation on cerebral oxygenation and intracranial pressure indicators, particularly in individuals with post-COVID-19 pneumonia, have not been adequately investigated. Therefore, our study sought to address this knowledge gap.

This study aimed to evaluate whether these primary and secondary outcomes differed between patients with and without a prior history of COVID-19 pneumonia and to compare subgroups receiving 2 L/min and 4 L/min supplemental oxygen in terms of all outcomes.

Materials and methods

After obtaining approval from the Non-Invasive Health Research Ethics Committee of Duzce University (Date: January 17, 2022; no: 2022/13), the study was registered at ClinicalTrials.gov (NCT05571683; retrospectively registered on October 7, 2022). This study was financially supported by the Duzce University Scientific Research Projects Coordination Office (DUBAP) under Project No: 2022.04.02.1326 and was conducted between May 2022 and January 2023, at the operating rooms of Duzce University Faculty of Medicine Health Practice and Research Center, including a total of 88 patients. This study was conducted in accordance with the Declaration of Helsinki. All patients signed a written informed consent form before participation. Because the COVID-19 history groups were observational and only oxygen-flow allocation was randomized, CONSORT reporting was applied specifically to the randomized oxygen-flow component within the two predefined COVID-19 history strata.

Patient selection

This prospective comparative study included patients undergoing cataract surgery with and without a history of COVID-19 pneumonia. The COVID-19 history groups were predefined according to patients’ medical history and were not randomized. Within each COVID-19 history stratum, patients were randomly assigned by block randomization (www.sealedenvelope.com) to receive either 2 L/min or 4 L/min oxygen insufflation under the surgical drape.

The study included 88 patients aged > 18 years, scheduled to undergo elective unilateral cataract surgery under local anesthesia using phacoemulsification and classified as American Society of Anesthesiologists (ASA) physical status I, II, or III. Eight patients were excluded from the study for various reasons and were not included in the statistical analyses. (Fig. 1).

Fig. 1.

Flow diagram of patient inclusion and stratified randomization

Patients who were anticipated to require postoperative intensive care unit admission, those scheduled for bilateral cataract surgery, and those with advanced chronic obstructive pulmonary disease, a history of coronary artery bypass surgery, or known hemoglobinopathies were excluded. Additionally, uncooperative patients, including those with Alzheimer’s disease, neurodegenerative disorders, or intellectual disability, were excluded from the study.

Study protocol

The study population consisted of 40 patients with (Group A) and 40 patients without (Group B) a history of COVID-19 pneumonia. One subgroup received 2 L/min oxygen supplementation (Groups A-2 and B-2), while the other received 4 L/min oxygen supplementation (Groups A-4 and B-4) (Fig. 1).

Before surgery, patients were evaluated in the preoperative assessment room, and data including age, gender, ASA physiological classification, medications, history of COVID-19 pneumonia and treatment location, SARS-CoV-2 vaccination status, and number of vaccine doses, were recorded.

For cataract surgery, local anesthesia was administered to the operative eye by the surgeon, and the patients were positioned supine on the operating table without a Trendelenburg tilt. ONSD was measured in the non-operated eye before application of the surgical drape and was reassessed at the end of surgery immediately after drape removal, while the patient remained in the supine position. All ONSD measurements were performed by a single investigator who was not masked to the patients’ COVID-19 history or oxygen-flow allocation, using a high-frequency, high-resolution linear ultrasound probe (4–13 MHz; MyLab X7, Esaote SpA, Genova, Italy). Measurements were obtained through the closed upper eyelid with the patient in the supine position. ONSD was measured 3 mm posterior to the point at which the optic nerve enters the globe. Three measurements were obtained at each assessment, and their mean value was recorded for analysis. Prior to sterilization of the operative eye, two sensors compatible with the Masimo O3 Root Monitor^® (Masimo Corporation, Irvine, CA, USA) were positioned on the right and left frontal sites for cerebral oximetry. Following device calibration and baseline determination, baseline measurements were obtained for cerebral oxygen saturation (rSO₂), relative deoxyhemoglobin (ΔHHb), oxyhemoglobin (ΔO₂Hb), and total hemoglobin (ΔcHb) changes in both temporofrontal regions. The device-derived ΔO₂Hb, ΔHHb, and ΔcHb values were referenced to the baseline established during O3 sensor initialization; the study time point defined as 0 min corresponded to the start of surgical draping, which occurred several minutes after device baseline establishment and ONSD measurement, and therefore the relative NIRS values at 0 min were not necessarily zero. Standard anesthesia monitoring was performed. ETCO₂ was measured with the Capnostream™20p device (Medtronic, Boulder, USA) through a dual-lumen side-stream nasal cannula with separate channels for oxygen delivery and exhaled CO₂ sampling (Smart CapnoLine™ Plus, Medtronic, Boulder, USA). The sampling flow rate of the device was 50 mL/min. ETCO₂ values were recorded only when a regular and interpretable capnographic waveform was obtained. If the waveform was irregular or inadequate, the sampling setup and device connections were checked and adjusted until an acceptable waveform was achieved. No predefined quantitative criteria regarding waveform amplitude or plateau characteristics were used. Once all baseline parameters were recorded, patients were handed over to the surgical team for the procedure.

According to their assigned subgroup, patients received supplemental oxygen at a flow rate of either 2 L/min or 4 L/min. Systolic arterial pressure (SAP), diastolic arterial pressure (DAP), mean arterial pressure (MAP), heart rate (HR), ETCO₂, integrated pulmonary index (IPI), respiratory rate (RR), and peripheral oxygen saturation (SpO₂) were measured and recorded every 5 min throughout the operation. Similarly, right and left rSO₂, ΔO₂Hb, ΔHHb, and ΔcHb were continuously monitored and recorded at 5-minute intervals. Total surgical duration and total drape duration were recorded separately for each participant. While the patient remained in the supine position on the operating table, the ONSD of the non-operated eye was re-measured ultrasonographically and recorded.

Statistical analysis

Based on an a priori power analysis for the repeated-measures main-group analysis of the prespecified primary outcome ETCO₂, assuming a moderate effect size of f = 0.25, a 5% significance level, 80% power, and a correlation of 0.50 between repeated measurements (PASS v.15 demo), 80 patients were required. A dropout rate of 10% was incorporated into the sample size projection. Of the 88 patients initially assessed for eligibility, 8 were excluded from the study, leaving 80 patients enrolled according to the study protocol and assigned to subgroups using block randomization (Fig. 1). The oxygen-flow subgroup analyses included 20 patients per subgroup and were not independently powered. The analytical strategy was structured around the primary comparison between patients with and without a history of COVID-19 pneumonia and the secondary comparisons within the predefined oxygen-flow strata. A formal COVID-19 history × oxygen-flow interaction was not estimated. Therefore, the oxygen-flow subgroup comparisons should be interpreted as exploratory.

The normality assumption for continuous variables was assessed using Shapiro–Wilk test, and homogeneity of variance was evaluated using Levene’s test. Sporadically missing values, which accounted for approximately 0.1% of otherwise available repeated measurements, were assumed to be missing completely at random (MCAR) and were handled using single mean or median imputation, as appropriate. Scheduled time points occurring after drape removal were considered structurally unavailable and were not imputed. Given the very small extent of sporadic missingness, this approach was retained, although single-value imputation may underestimate variance and affect standard errors.

Between-group comparisons were performed using Independent Samples t-test or Mann–Whitney U test, as appropriate. In some within-group comparisons, a paired-samples t-test was applied, obtaining the corrected p-value. To control the family wise error rate (FWER) in comparisons involving variables measured at multiple time points, the Generalized Estimating Equations (GEE) approach (post hoc: Bonferroni multiple comparison test) was used to obtain robust parameter estimates using an exchangeable correlation matrix structure and gamma with log link or normal with identity as link functions. Additionally, Bonferroni adjustment was applied across parameters to control the family-wise error rate for the group-by-time interaction tests in the main-group comparisons. Categorical variables were compared using the Pearson chi-square or Fisher–Freeman–Halton test, with Bonferroni post hoc adjustment when appropriate. Statistical significance was set at p < 0.05.

Results

No statistically significant differences were observed between the main groups and subgroups in terms of age, gender, ASA score, vaccination status, or drape and surgery duration (Table 1). Among patients in Group A with a history of COVID-19, no significant difference was found between subgroups A-2 and A-4 regarding the location of COVID-19 treatment (p = 0.797) (Supplementary Figure S1).

Table 1.

Comparison of groups and subgroups in terms of demographic data, ASA scores, and durations of surgery and drapes

		A (n = 40)	B (n = 40)		p	A-2 (n = 20)	B-2 (n = 20)	p	A-4 (n = 20)	B-4 (n = 20)	p
Gender	Male	20	19		0.823	9	10	0.752	11	9	0.527
	Female	20	21			11	10		9	11
Age (years)		66.1 ± 8.2		68.6 ± 11.1	0.262	66 ± 7.2	68 ± 12.4	0.537	66 ± 9.3	68.2 ± 10	0.341
ASA score		2.5[1]		2[1]	0.510	2[1]	2.5[1]	0.718	3[1]	2[1]	0.242
Surgery duration (min)		30 [4]		29 [9.5]	0.155	29.5[4.5]	26[14.8]	0.121	30[8.5]	29[3]	0.512
Drape duration (min)		36 [4.8]		36 [11.3]	0.199	36 [4]	32.5[15.8]	0.231	37 [9]	36 [3.8]	0.221
SARS-COV-2 Vaccine	Not vaccinated	0	1		0.999	0	1	0.999	0	0	0.744
	Incomplete	24	24			12	11		12	13
	Complete	16	15			8	8		8	7

Data are presented as n, mean ± standard deviation, or median [interquartile range], as appropriate. A: Group with COVID-19 history, B: Group without COVID-19 history, A-2: Group with COVID-19 history, receiving 2 L/min oxygen. B-2: Group without COVID-19 history, receiving 2 L/min oxygen. A-4: Group with COVID-19 history, receiving 4 L/min oxygen. B-4: Group without COVID-19 history, receiving 4 L/min oxygen

End-tidal carbon dioxide (ETCO₂) and other capnographic findings

No statistically significant time-dependent changes in ETCO₂ were observed within the main groups. Additionally, no significant difference was detected in ETCO₂ values between Groups A and B (p = 0.600). In subgroups A-2 and B-2, differences in ETCO₂ values measured at various time points were not statistically significant, either within or between the groups (p = 0.530). Similarly, in patients included in subgroups A-4 and B-4, differences in ETCO₂ values across different time points were not statistically significant within or between the groups (p = 0.958) (Table 2; Fig. 2).

Table 2.

Comparison of end-tidal carbon dioxide values between main groups and subgroups

Time (min)	GROUP A (n = 40)	GROUP B (n = 40)	p	GROUP A-2 (n = 20)	GROUP B-2 (n = 20)	p	GROUP A-4 (n = 20)	GROUP B-4 (n = 20)	p
0	34.5 ± 2.9	33.4 ± 4.1	0.600	34.3 ± 2.7	33.5 ± 3.6	0.530	34.7 ± 3.1	33.6 ± 4.7	0.958
5	35.0 ± 3.1	33.9 ± 4.6		34.9 ± 2.8	33.8 ± 4.0		35.1 ± 3.5	33.9 ± 5.3
10	35.2 ± 3.0	34.5 ± 4.2		35.2 ± 2.9	34.6 ± 3.5		35.2 ± 3.2	34.4 ± 4.9
15	35.5 ± 3.2	34.5 ± 4.2		35.8 ± 3.0	34.9 ± 3.8		35.3 ± 3.4	34.2 ± 4.6
20	35.7 ± 3.3	34.4 ± 3.8		35.9 ± 3.1	34.7 ± 3.4		35.6 ± 3.5	34.0 ± 4.1
25	35.9 ± 2.7	34.6 ± 3.7		36.2 ± 2.3	35.2 ± 3.5		35.6 ± 3.0	34.1 ± 3.9
30	36.1 ± 2.7	34.5 ± 3.8		36.0 ± 2.6	34.0 ± 2.7		36.3 ± 2.9	34.9 ± 4.3
35	36.8 ± 2.3	34.4 ± 3.6		37.0 ± 1.9	34.0 ± 2.9		36.6 ± 2.6	34.8 ± 4.3

Data are presented as mean ± standard deviation. Group sample sizes shown in the column headings indicate the total enrolled sample; the number of available observations decreased at later time points because measurements scheduled after drape removal were structurally unavailable and were not imputed. A: Group with COVID-19 history, B: Group without COVID-19 history, A-2: Group with COVID-19 history, receiving 2 L/min oxygen. B-2: Group without COVID-19 history, receiving 2 L/min oxygen. A-4: Group with COVID-19 history, receiving 4 L/min oxygen. B-4: Group without COVID-19 history, receiving 4 L/min oxygen

Fig. 2.

Graphical representation of end-tidal carbon dioxide (ETCO₂) levels recorded throughout the operation in the main groups

No statistically significant time-dependent changes in the Integrated Pulmonary Index (IPI) were observed within the main groups. Furthermore, no significant difference was found when comparing the measured IPI values between Groups A and B (p = 0.587). In subgroups A-2 and B-2, differences in IPI values measured at various time points were not statistically significant, either within or between the groups (p = 0.522). Similarly, in patients included in subgroups A-4 and B-4, differences in IPI values across time points were not statistically significant within or between the groups (p = 0.231) (Supplementary Table S1).

No statistically significant time-dependent changes in respiratory rate (RR) were observed within the main groups. Furthermore, no significant difference was detected when comparing the measured RR values between Groups A and B (p = 0.231). In subgroups A-2 and B-2, differences in RR values measured at various time points were not statistically significant, either within or between the groups (p = 0.060). Similarly, in subgroups A-4 and B-4, differences in RR values across time points were not statistically significant within or between the groups (p = 0.467) (Supplementary Table S1).

Cerebral oximetric findings

Right and left frontal cerebral oxygen saturation (rSO₂) values showed no significant group-by-time differences between Groups A and B (p = 0.718 and p = 0.597, respectively). Similarly, in subgroups receiving 2 L/min or 4 L/min oxygen supplementation, rSO₂ values remained stable over time and did not differ significantly within or between groups. The relative changes in total hemoglobin (ΔcHb) were also comparable between groups for both hemispheres (p = 0.421 for right, p = 0.754 for left). Subgroup analyses confirmed the absence of significant intra- or intergroup differences. Left frontal relative deoxyhemoglobin (ΔHHb-left) and right and left oxyhemoglobin changes (ΔO₂Hb) likewise showed no significant differences between groups or across time (p = 0.120, p = 0.860, and p = 0.155, respectively). Corresponding subgroup analyses for both oxygen supplementation levels did not reveal any significant changes (Supplementary Table S2).

In contrast, right frontal deoxyhemoglobin changes (ΔHHb-right) demonstrated significant group-by-time interaction between Groups A and B (p = 0.002). At the 10th minute, ΔHHb-right values differed significantly between Groups A and B (p = 0.003), and within Group A, ΔHHb-right values changed significantly from baseline to the 10th minute (p = 0.045). Significant group-by-time interactions were also observed between Groups A-2 and B-2 (p = 0.003) and between Groups A-4 and B-4 (p = 0.006). In the subgroup analysis, ΔHHb-right values differed significantly between Groups A-2 and B-2 at the 35th minute (p < 0.001) and between Groups A-4 and B-4 at the 10th minute (p = 0.004). However, no consistent intergroup differences were observed when the oxygen supplementation subgroups were directly compared (Table 3).

Table 3.

Right frontal deoxyhemoglobin changes (ΔHHb-right) in the main group and subgroup analyses

Time (min)	GROUP A (n = 40)	GROUP B (n = 40)	p	GROUP A-2 (n = 20)	GROUP B-2 (n = 20)	p	GROUP A-4 (n = 20)	GROUP B-4 (n = 20)	p
0	-0.04 ± 1.67	-0.04 ± 1.65	0.002	-0.02 ± 1.62	-0.25 ± 1.56	0.003	-0.05 ± 1.75	0.17 ± 1.75	0.006
5	-0.10 ± 3.35	-1.32 ± 2.16		-0.35 ± 2.03	-1.06 ± 2.21		0.13 ± 4.30	-1.58 ± 2.14
10	-0.92 ± 3.70	-1.76 ± 2.85		-1.87 ± 2.07	-1.51 ± 3.21		-0.02 ± 4.64	-2.00 ± 2.50
15	-1.88 ± 3.49	-2.39 ± 3.11		-2.58 ± 2.49	-2.51 ± 3.78		-1.22 ± 4.19	-2.26 ± 2.36
20	-2.23 ± 3.23	-2.20 ± 2.72		-2.84 ± 2.42	-2.00 ± 3.07		-1.64 ± 3.82	-2.40 ± 2.37
25	-2.39 ± 3.76	-2.16 ± 3.27		-3.01 ± 2.97	-1.79 ± 4.02		-1.79 ± 4.38	-2.52 ± 2.36
30	-2.49 ± 3.53	-2.51 ± 2.85		-2.84 ± 2.60	-2.12 ± 3.78		-2.16 ± 4.28	-2.72 ± 2.26
35	-2.26 ± 4.17	-2.64 ± 3.02		-2.89 ± 3.19	-2.20 ± 3.53		-1.83 ± 4.82	-3.08 ± 2.52

Data are presented as mean ± standard deviation.The p values represent the overall group-by-time interaction effects obtained from the GEE models. Group sample sizes shown in the column headings indicate the total enrolled sample; the number of available observations decreased at later time points because measurements scheduled after drape removal were structurally unavailable and were not imputed. A: Group with COVID-19 history, B: Group without COVID-19 history, A-2: Group with COVID-19 history, receiving 2 L/min oxygen. B-2: Group without COVID-19 history, receiving 2 L/min oxygen. A-4: Group with COVID-19 history, receiving 4 L/min oxygen. B-4: Group without COVID-19 history, receiving 4 L/min oxygen. Bold values indicate statistically significant p values (p<0.05)

In summary, while most cerebral oximetry parameters, including rSO₂, ΔcHb, ΔHHb-left, and ΔO₂Hb, remained stable across groups and time points, right frontal deoxyhemoglobin (ΔHHb-right) exhibited significant group- and time-dependent differences.

Optic nerve sheath diameter

No statistically significant differences were observed between Group A and Group B in pre- and postoperative ONSD measurements (p = 0.096). Similar findings were obtained when comparing patients in subgroups A-2 and B-2 (p = 0.192) and in subgroups A-4 and B-4 (p = 0.302) (Table 4; Fig. 3). Paired within-group analyses demonstrated that postoperative ONSD values were significantly higher than preoperative values in Group A (p < 0.001, Cohen’s d effect size = -0.79, 95% CI: -1.14 to -0.43) and Group B (p < 0.001, Cohen’s d effect size = -0.76, 95% CI: -1.10 to -0.40). Also, the pre- and postoperative ONSD measurements were significantly different in Group A-2 (p = 0.034, Cohen’s d effect size = -0.70, 95% CI: -1.18 to -0.20), Group B-2 (p = 0.048, Cohen’s d effect size = -0.65, 95% CI: -1.13 to -0.16), Group A-4 (p = 0.005, Cohen’s d effect size = -0.89, 95% CI: -1.40 to -0.46), Group B-4 (p = 0.008, Cohen’s d effect size = -0.85, 95% CI: -1.35 to -0.32).

Table 4.

Comparison of optic nerve sheath diameter values between the main groups and subgroups

	GROUP A (n = 40)	GROUP B (n = 40)	p	GROUP A-2 (n = 20)	GROUP B-2 (n = 20)	p	GROUP A-4 (n = 20)	GROUP B-4 (n = 20)	p
Baseline (mm)	5.45 ± 0.60	5.65 ± 0.66	0.096	5.42 ± 0.59	5.74 ± 0.67	0.192	5.49 ± 0.62	5.56 ± 0.65	0.302
End of surgery (mm)	5.83 ± 0.46	5.89 ± 0.77		5.78 ± 0.47	5.95 ± 0.82		5.89 ± 0.45	5.83 ± 0.73

Data are presented as mean ± standard deviation. A: Group with COVID-19 history, B: Group without COVID-19 history, A-2: Group with COVID-19 history, receiving 2 L/min oxygen. B-2: Group without COVID-19 history, receiving 2 L/min oxygen. A-4: Group with COVID-19 history, receiving 4 L/min oxygen. B-4: Group without COVID-19 history, receiving 4 L/min oxygen

Fig. 3.

Preoperative and postoperative optic nerve sheath diameter (ONSD) measurements in the main groups

Hemodynamic findings

No statistically significant differences were observed within or between main groups regarding SAP, DAP, MAP, HR. However, a significant group-by-time interaction was observed for DAP between the 2 L/min subgroups (p = 0.002). A separate baseline comparison showed no statistically significant difference in DAP between Groups A-2 and B-2 (p = 0.214).

Regarding time-dependent changes in SpO₂, no significant differences were found between Groups A and B (p = 0.999); however, significant time-dependent changes in SpO₂ were observed within both the main groups and subgroups. In all groups, SpO₂ measured at baseline was significantly lower than values measured at subsequent time points (p < 0.001). Although some time-point comparisons revealed statistically significant differences in SpO₂ values between groups (p = 0.003–0.030), group mean SpO₂ values remained above 95%; no SpO₂ value below 90% was recorded, and no clinically relevant desaturation occurred (Supplementary Tables S3, S4).

Discussion

In phacoemulsification surgery performed under local anesthesia, the operative field is covered with a sterile surgical drape to retract eyelashes from the surgical site. Due to the air-impermeable nature of these drapes, CO₂ accumulation during the procedure has been reported in several studies [4–7]. The adverse effects of CO₂ accumulation under the surgical drape on respiratory physiology during ophthalmic surgery have been previously described. In a study by Schlager in 1999, it was reported that in patients without respiratory disorders and not receiving sedation, ETCO₂ levels significantly increased under the surgical drape. Although oxygen insufflation at 2 L/min protected against hypoxia, it did not prevent hypercapnia [5]. These findings are further supported by Schlager and Luger (2000), who demonstrated that fixed-flow oxygen delivered via a nasal cannula prevented hypoxia but did not eliminate rebreathing of CO₂ under the surgical drape [6]. Similarly, in a study by Risdall and Geraghty on phacoemulsification surgery under local anesthesia, oxygen supplementation at 2 L/min via nasal cannula was observed to prevent hypoxia, leading to the recommendation that all patients undergoing ophthalmic surgery under local anesthesia should routinely receive oxygen support [11]. In a randomized controlled study by Sukcharanjit et al. involving a similar patient population undergoing ocular surgery under local anesthesia, the effects of forced air heaters were compared with those of controlled heating overhead blankets. The study demonstrated that forced air heaters facilitated air movement under the drape, promoting CO₂ washout and preventing CO₂ accumulation, whereas such effects were not observed with blankets that did not provide airflow [4]. Zeitlin et al. found that oxygen insufflation administered under a sterile drape at a flow rate at least twice the patient’s minute ventilation prevented increases in CO₂ levels during cataract surgery [7].

In hypercapnic patients, hyperventilation and increased cerebral blood flow serve as physiological compensatory mechanisms [8]. Hypercapnia, and the resulting increase in cerebral blood flow, is known to cause a clinically relevant increase in intracranial pressure and, consequently, in ONSD [9, 10].

COVID-19, which primarily affects the lungs but can involve multiple organs, has continued to manifest as Post-COVID-19 condition and requires comprehensive investigation due to its long-term effects [12, 13]. Based on the literature, our study hypothesis was that CO₂ accumulation under the surgical drape would increase ETCO₂ levels in all patient groups, leading to cerebral vasodilatation and, consequently, an increase in cerebral oxy- and deoxyhemoglobin changes, and ONSD measurements. Our primary aim was to investigate whether these outcomes differed between patients with and without a history of COVID-19. Our secondary aim was to compare all outcomes between subgroups receiving oxygen supplementation at 2 L/min and 4 L/min.

The volunteers included in our study were comparable in terms of age, gender, ASA score, drape and surgery duration, and COVID-19 vaccination status across the main and subgroups. No significant differences were observed between the subgroups of patients in Group A in terms of the location of treatment. Therefore, all main groups and subgroups were similar in terms of demographic characteristics, and no differences were noted in disease severity among patients with a history of COVID-19.

In our study, no statistically significant differences in ETCO₂ levels were observed between the main groups or subgroups throughout the operation. Increased ETCO₂ levels due to CO₂ accumulation under the sterile drape during phacoemulsification under local anesthesia have been reported in previous studies [4, 14]. In a study by Mumoli et al. on the respiratory function of post-COVID-19 patients, arterial blood gases were evaluated three months after hospitalization for COVID-19, and it was reported that 40% of the patients were hypocapnic [15]. In contrast to this study, in our research, ETCO₂ levels at baseline in patients with a history of COVID-19 were within normal limits; however, the absence of arterial blood gas analysis in our study represents a limitation. Furthermore, graphical representation of ETCO₂ levels shows that in Group A, consisting of patients with prior COVID-19, ETCO₂ values were approximately 1 mmHg higher throughout the operation than those in Group B, contrary to the expected hypocapnia reported in the referenced study. Nevertheless, this difference was neither clinically nor statistically significant. (Fig. 2)

In the present study, ETCO₂ was measured using a dual-lumen nasal cannula with separate channels for oxygen delivery and CO₂ sampling, thereby minimizing direct admixture of supplemental oxygen into the sampling line. Nevertheless, in spontaneously breathing, non-intubated patients, oxygen delivered through the nasal route may still mix with exhaled gas around the nares and result in some underestimation of true alveolar CO₂. This effect may be more pronounced at higher oxygen flow rates. Because simultaneous arterial or transcutaneous CO₂ measurements were not obtained, the magnitude of this potential measurement bias could not be quantified.

IPI is a value provided by the Capnostream™20p device that combines RR, HR, SpO₂, and ETCO₂ through an algorithmic system, presenting the patient’s respiratory function during monitoring as a score ranging from 1 to 10 [16]. In our study, no statistically significant differences in IPI scores were observed between the main groups or subgroups. During monitoring, no respiratory complications occurred in any patient, and no interventions were required due to decreases in IPI scores. No statistically significant differences in RR were observed between the main groups or subgroups. However, the time-dependent comparison of RR in the A-2 and B-2 subgroups receiving 2 L/min oxygen revealed a p-value of 0.060, which drew our attention. The mean RR of patients who recovered from COVID-19 in the A-2 group were consistently higher than those in the B-2 group at all measured time points; however, this difference was not statistically significant. When ETCO₂, IPI scores, and RR were evaluated together, the discrepancy between our findings and those of previous studies may be related to differences in study populations, monitoring methods, oxygen supplementation protocols, and the relatively short duration of surgery in our cohort, which averaged approximately 30 min.

In our study, the rSO₂, ΔcHb, ΔO₂Hb, and ΔHHb obtained from right and left temporofrontal regions were compared between the main and subgroup populations. According to the results, there were no significant differences between the groups in terms of bilateral rSO₂, ΔcHb, or ΔO₂Hb values. Similarly, ΔHHb-left values showed no significant difference. However, ΔHHb-right values differed significantly between Groups A and B at the 10th minute. Within Group A, ΔHHb-right values changed significantly from baseline to the 10th minute. In the subgroup analysis, significant differences were observed between Groups A-2 and B-2 at the 35th minute and between Groups A-4 and B-4 at the 10th minute.

The rSO₂ values obtained using near-infrared spectroscopy (NIRS) technique represent regional tissue perfusion and provide high accuracy in evaluating changes in cerebral blood flow. In addition, the ΔcHb, ΔO₂Hb, and ΔHHb parameters offer complementary information useful for monitoring the cerebral hemodynamic status [17]. Because this method does not measure arterial flow alone, it operates independently of pulsatility. The measurements reflect approximately 70% venous and 30% arterial contribution, and normal rSO₂ values range between 60–75% [18]. As expected, an increase in cerebral blood flow is accompanied by a rise in ΔO₂Hb, whereas conditions that impair venous return or cause intracranial congestion lead to an increase in ΔHHb [19].

In our study, rSO₂ values were within the normal range in all participants, and no cases of hypoxia or mental deterioration were observed. Although significant differences were detected in the ΔHHb-right, no statistical differences were found in ΔcHb, or in ΔO₂Hb. Moreover, the observed differences were below 10%, remaining well below the threshold of clinical significance accepted for trend monitoring [20, 21]. Therefore, the unilateral difference in ΔHHb-right does not appear to be clinically significant.

In contrast, several studies using the NIRS technique have demonstrated that prefrontal cortex activities (such as verbal fluency, mental arithmetic, driving simulation, speech, and language processing) as well as ocular movements and accommodation stimuli can induce changes in ΔcHb, ΔO₂Hb, and ΔHHb levels [22–27]. The participants in our study were surgical patients who were not sedated; therefore, environmental stimuli during the operation, such as staff conversations, verbal interactions with the patient, spontaneous eye movements, background music, or surgical light brightness, may have been associated with the statistically significant differences observed in ΔHHb-right values. This represents an important limitation of our study and a finding that warrants further research.

No statistically significant differences were observed between the main groups or corresponding oxygen-flow subgroups in the between-group comparisons of ONSD. In contrast, paired within-group analyses demonstrated that postoperative ONSD values were significantly higher than preoperative values in both main groups and all four oxygen-flow subgroups. Because ONSD was measured only before application of the surgical drape and immediately after drape removal, this finding should be interpreted as a pre-to-postoperative increase rather than as a gradual intraoperative change. Although the differences were statistically significant, their clinical relevance remains uncertain, as no consistent between-group difference or neurological deterioration was observed. ONSD provides valuable noninvasive information regarding changes in intracranial pressure, as the anatomy of the optic nerve sheath allows it to reflect variations in cerebrospinal fluid pressure [9, 10, 28, 29]. In our study, although statistical analyses of ETCO₂ values measured at regular intervals between the beginning and end of the operation did not reveal significant differences between groups, an increase in ETCO₂ levels was observed within each group throughout the procedure (Fig. 2). The duration of the surgical procedures may have limited the extent of physiological changes associated with drape-covered surgery.

The surgical durations observed in the present study should also be interpreted in the context of the university teaching-hospital setting, where cataract surgeries were generally performed by ophthalmology residents in their second through fourth years of training under the direct supervision of senior ophthalmic surgeons. This setting may partly account for the longer surgical durations compared with procedures performed entirely by experienced senior surgeons (Supplementary Table S5). Therefore, our findings should be interpreted within the context of unilateral phacoemulsification performed under local anesthesia. Further studies involving longer procedures and different patient populations are required to determine whether these findings remain consistent across different clinical conditions.

When hemodynamic parameters were analyzed, no significant differences were found between the main groups in SAP, DAP, MAP, or in HR. However, DAP differed significantly between the subgroups receiving 2 L/min oxygen (p = 0.002), being higher in the A-2 subgroup with COVID-19 history. Among all participants, 29 of 40 patients without and 20 of 40 patients with prior COVID-19 had hypertension. A review on the cardiovascular effects of COVID-19 reported a strong link between hypertension and SARS-CoV-2 infection, possibly due to persistent inflammation and cytokine activation. In addition, the virus’s affinity for ACE-2 receptors may limit the use of ACE inhibitors or angiotensin receptor blockers [30]. The combined effects of post-COVID inflammation, preexisting hypertension, and the absence of sedoanalgesia may explain the observed rise in DAP. In a study by Yorulmaz et al. on ocular surgeries under local anesthesia, increases in ETCO₂ under surgical drapes were not accompanied by changes in MAP, HR, arrhythmias, or SpO₂ [14]. Similarly, in our study, the higher DAP did not cause differences in MAP between groups, suggesting a lack of clinical relevance.

No significant differences were found between groups in time-dependent SpO₂ changes. Within-group analyses showed lower baseline values compared with later measurements. As baseline SpO₂ levels, recorded before oxygen supplementation, remained within physiological limits, no intergroup differences were observed, and no hypoxic events occurred, this statistically significant change was not clinically meaningful.

In conclusion, in terms of ETCO₂ and ONSD, no statistically or clinically significant differences were observed between groups. In subgroups with COVID-19 history, ΔHHb-right values differed significantly from those in the corresponding subgroups in Group B. Overall, the limited clinical significance of these findings may be related to the relatively short duration of the cataract surgeries. Further studies involving different ophthalmic procedures and longer operative durations are required to determine whether similar findings are observed under different clinical conditions. Because all participants received supplemental oxygen and no unsupplemented control group was included, the present study cannot determine the clinical benefit of oxygen supplementation itself. Recommendations regarding routine supplemental oxygen and appropriate intraoperative respiratory monitoring during ophthalmic surgery with impermeable drapes are therefore based on previous literature and established clinical considerations rather than on a direct comparison in the present study.

Limitations

Our study has several limitations. First, arterial blood gas analysis was not performed, and the possibility of hypocapnia reported in the literature was therefore not evaluated. Although a dual-lumen nasal cannula with separate oxygen-delivery and CO₂-sampling channels was used, nasal oxygen may still have diluted exhaled gas at the sampling site, potentially leading to underestimation of true alveolar CO₂, particularly at 4 L/min. The absence of simultaneous arterial or transcutaneous CO₂ measurements precluded quantification of this potential bias.

ONSD ultrasonography is operator-dependent and subject to measurement variability. All measurements were performed by a single investigator who was not masked to group assignment, which may have introduced measurement bias. Although a standardized technique was used and three consecutive measurements were averaged, the individual readings were not retained separately; therefore, intra-observer reproducibility could not be assessed retrospectively. Inter-observer reproducibility was also not evaluated because measurements were performed by only one investigator.

Although multiplicity across parameters was additionally assessed, the oxygen-flow subgroup analyses were not independently powered. Therefore, the isolated significant findings observed in the 2 L/min subgroup comparisons should be regarded as exploratory and hypothesis-generating.

An additional limitation is that, due to the observational nature of our study, no effort or intervention was made to minimize various stimuli to which the volunteers were exposed during surgery, which could potentially affect the prefrontal cortex and influence cerebral oximetry parameters. These stimuli were neither recorded nor analyzed.

Another limitation is that unilateral phacoemulsification surgeries under local anesthesia at our institution were completed in a relatively short duration, averaging approximately 30 min, which may not have allowed sufficient time for CO₂ accumulation to reach levels that could produce significant differences. Therefore, caution should be exercised when extrapolating these results to longer ophthalmic procedures, different anesthetic techniques, or patients with significant respiratory impairment.

Finally, although ASA physical status was reported as a baseline characteristic, the study was not designed or powered to perform subgroup analyses according to the ASA classification. Because ASA status was not used as a stratification factor during randomization and the number of patients within individual ASA categories would be limited, any post hoc ASA-based subgroup analysis would be exploratory and potentially underpowered. Therefore, no ASA physical status-stratified analysis was conducted.

Conclusion

In patients undergoing short-duration unilateral phacoemulsification under local anesthesia with supplemental oxygen at 2–4 L/min, ophthalmic draping was not associated with clinically meaningful changes in ETCO₂ or ONSD, regardless of the history of COVID-19 pneumonia. Although selected cerebral oximetry parameters showed statistically significant differences between groups, all measured respiratory, hemodynamic, and cerebral physiological variables remained within physiological limits throughout the procedure. These findings should be interpreted within the context of the specific clinical setting studied and should not be generalized to longer ophthalmic procedures or patients with significant respiratory compromise.

Abbreviations

CO₂

Carbon dioxide

COVID-19

Coronavirus disease 2019

SAP

Systolic arterial pressure

DAP

Diastolic arterial pressure

MAP

Mean arterial pressure

HR

Heart rate

ETCO₂

End-tidal carbon dioxide

PaCO₂

Partial pressure of arterial carbon dioxide

IPI

Integrated pulmonary index

RR

Respiratory rate

SpO₂

Peripheral oxygen saturation

ONSD

Optic nerve sheath diameter

rSO₂

Regional/cerebral oxygen saturation

ΔO₂Hb

Relative oxyhemoglobin change

ΔHHb

Relative deoxyhemoglobin change

ΔcHb

Relative total hemoglobin change

NIRS

Near-infrared spectroscopy

Authors' contributions

KO: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Visualization, Writing – original draft, Writing – review & editing, Final approval of manuscript. OEK: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Visualization, Writing – original draft, Writing – review & editing, Final approval of manuscript. SC : Formal analysis, Methodology, Resources, Validation, Writing – review & editing, Final approval of manuscript. GDS: Data curation, Resources, Writing – review & editing, Final approval of manuscript. ISY: Data curation, Resources, Writing – review & editing, Final approval of manuscript. GS: Data curation, Resources, Supervision, Writing – review & editing, Final approval of manuscript. YD: Data curation, Resources, Supervision, Writing – review & editing, Final approval of manuscript.

Funding

This study was financially supported by the Duzce University Scientific Projects Coordination Office (DUBAP) under Project Number 2022.04.02.1326.

Data availability

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

Duzce University Non-Invasive Health Research Ethics Committee; Date: January 17, 2022; no: 2022/13. All patients signed a written informed consent form before participation.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.