Effectiveness of personalized positioning management in preventing hypoxemia during sedated endoscopy: a retrospective study

Abstract

Background

To investigate the preventive effects of personalized positioning management on hypoxemia during sedated endoscopic examination in elderly patients with esophageal cancer (EC).

Methods

A retrospective observational study was conducted. We retrospectively reviewed the medical records of elderly patients with esophageal cancer who underwent sedated gastroscopy between January 2023 and June 2024. Patients who received personalized positioning management were assigned to the intervention group (n = 67), while those who underwent conventional left lateral decubitus positioning were assigned to the control group (n = 67). The incidence of hypoxemia, lowest peripheral oxygen saturation (SpO₂), severe hypoxemic events, examination time, sedative dosage, comfort scores, and incidence of adverse events were compared between the two groups. Binary logistic regression analysis was performed to identify factors associated with hypoxemia.

Results

In the intervention group, the incidence of hypoxemia was significantly lower; the lowest SpO₂ was distinctly higher; and the incidences of severe hypoxemia and hypoxemic events requiring intervention were lower. Personalized positioning management was identified as an independent protective factor against hypoxemia. A higher American Society of Anesthesiologists physical status classification (Class III) was an independent risk factor for hypoxemia. The intervention group had a shorter examination time, lower propofol dosage, higher comfort scores, and a lower incidence of adverse events.

Conclusion

Personalized positioning management effectively reduces both the incidence and severity of hypoxemia during sedated endoscopic examination in elderly patients with EC. This strategy demonstrates high clinical safety and feasibility.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12876-026-05168-8.

Introduction

Esophageal cancer (EC) is a malignant tumor of the digestive tract worldwide. More than 640,000 EC cases and near 590,000 related deaths have been reported globally in 2022. In China, the age-standardized incidence rate is approximately 13.7 per 100,000 population, indicating a particularly heavy disease burden [1]. With the accelerating process of population aging, patients aged 65 years and older account for more than 50% of newly diagnosed cases of EC [2]. Under these circumstances, endoscopic examination is crucial in the early screening, diagnosis, and pathological staging of EC in clinical practice.

However, elderly patients frequently exhibit age-related deterioration in respiratory physiological function when undergoing sedated endoscopic examination [3]. Previous studies have shown that the incidence of hypoxemia after sedation in elderly patients can reach 25%–40%, which is significantly higher than that observed in younger populations [4, 5]. Multiple clinical observations have reported that approximately 15%–30% of elderly patients examined in the conventional position experience hypoxemic events requiring clinical intervention. Moreover, in elderly patients with EC who have comorbid conditions such as obesity and chronic obstructive pulmonary disease (COPD), airway management during sedation is particularly challenging and often suboptimal [6, 7].

Personalized positioning management is based on principles of respiratory physiology and involves a comprehensive assessment of factors such as thoracic cage mobility, diaphragmatic motion, and airway patency to design individualized positioning strategies for each patient. Previous studies have demonstrated that appropriate positional adjustments can increase functional residual capacity by approximately 10%–15%, thereby effectively improving oxygenation status [8].

Although the fundamental concept of positioning management has been widely accepted, systematic investigations focusing specifically on elderly patients with EC, a high-risk population for peri-procedural hypoxemia, remain limited. EC represent a distinct high-risk subgroup for hypoxemia during sedated endoscopy, due to a convergence of age-related physiological decline and tumor-specific hazards that are synergistically amplified in this population. Ageing diminishes respiratory reserve through reduced chest wall compliance, muscle strength, and hypoxic ventilatory response [9]. This decline is compounded by tumor-specific risks: esophageal tumors can directly compress the trachea or main bronchi, tumor-related luminal obstruction elevates aspiration risk [10], and dysphagia-related malnutrition is notably prevalent. A recent review reported that sarcopenia affects 46.3% ± 19.6% of EC patients [11], and this malnutrition-driven muscle loss directly impairs respiratory function—as documented in a 72-year-old EC patient (BMI 19.4 kg/m²) whose maximal inspiratory pressure was only 23.6% of the predicted value [12]. Sugawara et al. further confirmed the clinical impact, showing that low %vital capacity (HR = 2.01) and low FEV1.0 (HR = 1.6) independently predicted non-esophageal cancer-related death in elderly EC patients [9]. These tumor-specific hazards are particularly detrimental in older patients, whose blunted protective reflexes and poor hypoxic tolerance render them far more vulnerable to transient airway compromise or minor aspiration than younger individuals. This synergy distinguishes elderly EC patients from other elderly cancer populations and justifies targeted positioning strategies. Therefore, this study retrospectively analyzed the clinical data to evaluate the preventive effects of personalized positioning management on hypoxemia during sedated endoscopic examination in elderly patients with EC. The findings are expected to provide evidence for the development of evidence-based, individualized positioning management protocols and to optimize peri-procedural safety strategies for the elderly.

Materials and methods

Study design

In this single-center retrospective observational study, medical records of elderly patients (≥ 65 years) with pathologically confirmed EC who underwent sedated gastroscopy between January 2023 and June 2024 in the Department of Gastroenterology of our hospital were retrospectively reviewed. All included patients underwent sedated gastroscopy for routine surveillance or diagnostic evaluation. Procedures involving therapeutic interventions (e.g., endoscopic resection, dilation, or stenting) were excluded from this study.

Participants

Inclusion criteria

Age ≥ 65 years; Pathologically confirmed diagnosis of EC; American Society of Anesthesiologists (ASA) physical status classification II–III [13]; Clear consciousness, basic communication ability, and capacity to cooperate with all examinations and assessments.

Exclusion criteria

Presence of severe cardiovascular or cerebrovascular diseases; Presence of severe respiratory system diseases; Known difficult airway or a high risk of aspiration; Body mass index (BMI) > 30 kg/m² [14]; History of allergy or adverse reactions to sedative or anesthetic agents; Inability to cooperate with positioning assessment.

Study procedures

Monitoring and oxygen protocol

All patients in both groups underwent sedated endoscopy under standardized monitoring, including electrocardiography, noninvasive blood pressure measurement, and pulse oximetry. All patients received supplemental oxygen via nasal cannula at 2 L/min throughout the procedure. The same monitoring and oxygen protocol was applied to both groups.

Position protocol

Based on medical record documentation, patients were categorized into two groups according to the positioning applied during sedated gastroscopy: conventional positioning or personalized positioning management.

Control group: patients were placed in the standard left lateral decubitus position as routinely performed in our center. Patients lay on their left side on the examination bed, with the head supported by a standard pillow (approximately 6–8 cm in height) to maintain approximate alignment of the cervical and thoracic spine. Both knees were naturally flexed at an angle of approximately 90°–120°. This position was maintained throughout the procedure, as documented in anesthesia or nursing records.

Intervention group: Personalized positioning management was implemented based on pre-procedural assessments documented in the medical records, following a standardized three-step protocol.

Step 1 - Assessment: The following parameters were extracted from pre-procedural nursing and anesthesia records [15]:

1. Thoracic mobility: assessed by the change in chest circumference (cm) during maximal inspiration and expiration.

2. Cervical spine range of motion: active flexion, extension, lateral flexion, and rotation (categorized as normal, mildly limited, or severely limited).

3. Tumor location (upper/mid/lower esophagus) and presence of dysphagia.

4. Comorbidities including COPD, asthma, or other respiratory conditions.

Step 2 - Individualized adjustments: Based on the assessment findings, the following modifications were applied:

1. Head elevation: 15°-30° using an adjustable pillow to maintain neutral cervical alignment.

2. Upper body elevation: 10°-20° in patients with severe dysphagia, proximal esophageal tumors, or COPD.

3. Shoulder support: additional padding to prevent neck hyperextension or excessive torsion.

4. The left lateral decubitus position was maintained as the base position in all patients.

Step 3 - Documentation: The final positioning parameters (head elevation angle, upper body elevation, use of shoulder support) were recorded in the nursing checklist before procedure initiation.

Adjustments were individualized; not all modifications were applied to every patient. The specific positioning strategy for each patient was determined by the attending anesthesiologist and endoscopy nurse based on the Step 1 assessment.

Propofol sedation was administered by experienced anesthesiologists using a standardized clinical protocol. The same sedation protocol was applied to both groups. Due to the retrospective nature of this study, quantitative sedation scores were not systematically recorded.

Positioning-related data were retrospectively extracted from standardized nursing checklists and anesthesia records, which routinely documented pre-procedural assessments (thoracic and cervical mobility, tumor location, comorbidities) and final positioning parameters (head elevation angle, upper body elevation, use of shoulder support). These records were completed by the attending anesthesiologist and endoscopy nurse before procedure initiation. Documentation of positioning was consistent across all included patients, as these items were part of the institutional sedation protocol checklist.

Outcome measures

Primary outcomes were extracted from the medical and anesthesia records, and included: (1) Incidence of hypoxemia: Defined as a decrease in SpO₂ below 90% lasting for more than 10 s during the examination [15]; (2) Lowest SpO₂: The minimum SpO₂ value recorded throughout the entire procedure; (3) Incidence of severe hypoxemia: Defined as any occurrence of SpO₂ below 85% [6]; (4) Incidence of hypoxemic events requiring intervention: Hypoxemic event requiring medical interventions such as airway adjustment or enhanced ventilatory support.

Secondary outcomes obtained from medical records included: (1) Examination time: Total procedural duration from endoscope insertion to withdrawal; (2) Total sedative dosage: Cumulative dose of sedative agents, including propofol; (3) Patient comfort assessment: Evaluated using the Visual Analog Scale (VAS), with scores ranging from 0 to 10 [16]; (4) Vital sign parameters: Changes in heart rate and blood pressure during the examination; (5) Demographic and clinical characteristics: Including age, sex, BMI, and ASA-PS.

Data collection

Demographic and clinical characteristics, procedural details, sedation dosage, intraoperative SpO₂ values, hypoxemic events, and interventions were retrospectively collected from the electronic medical record system, anesthesia charts, and nursing records.

SpO₂ was continuously monitored using a multi-parameter vital signs monitor, and the lowest SpO₂ value as well as hypoxemic events were recorded in real time by the study nurse. Examination time was documented by the endoscopist. Propofol dosage was recorded by the anesthesiologist. VAS comfort scores were assessed by the recovery room nurse after patient awakening.

All relevant data were extracted using a standardized data collection form designed for this study. For quality control, data entry was independently verified by two researchers. The final anonymized dataset was then used for statistical analysis.

Statistical analysis

With application of SPSS version 26.0 (IBM Corp., Armonk, NY, USA), the Shapiro-Wilk test was applied to assess the normality of measurement data. Normally distributed measurement data are presented as mean ± standard deviation (x̄±s) and were compared between groups using the independent-samples t test. Non-normally distributed measurement data are presented as median (interquartile range) [M (IQR)] and were compared using the Mann–Whitney U test. Enumeration data are expressed as number (percentage) [n (%)] and were compared using the χ² test or Fisher’s exact test. Binary logistic regression analysis was conducted to identify independent factors associated with the occurrence of hypoxemia, with odds ratios (ORs) and corresponding 95% confidence intervals (CI) calculated. All statistical tests were two-sided, and a P value < 0.05 was considered statistically significant.

Results

Comparison of baseline characteristics between the two groups

A total of 134 elderly patients who underwent sedated gastroscopy were retrospectively included in this study. Among them, 67 patients had received personalized positioning management and were categorized as the intervention group, while 67 patients had received conventional positioning and were categorized as the control group. There were no statistically significant differences between the two groups in baseline characteristics (all P > 0.05), indicating good comparability between the groups (Table 1).

Table 1.

Comparison of baseline characteristics between the two groups

Variables	Intervention group (n = 67)	Control group (n = 67)	t/χ²	P
Age (years)	72.4 ± 5.1	71.8 ± 5.6	t = 0.658	0.512
Sex, n(%)			χ²=0.121	0.728
Male	42 (62.7)	40 (59.7)
Female	25 (37.3)	27 (40.3)
BMI (kg/m²)	23.5 ± 2.8	23.2 ± 3.1	t = 0.598	0.551
ASA-PS classification, n(%)			χ²=0.314	0.855
Class Ⅱ	45 (67.2)	43 (64.2)
Class Ⅲ	22 (32.8)	24 (35.8)
Comorbid COPD, n(%)	11 (16.4)	9 (13.4)	χ²=0.241	0.624
Smoking history, n(%)	35 (52.2)	38 (56.7)	χ²=0.278	0.598

Comparison of hypoxemia-related outcomes

The incidence of hypoxemia among patients who had received personalized positioning management was 11.9% (8/67), which was significantly lower than that among patients who had received conventional positioning (37.3%, 25/67) (P < 0.05). The mean lowest SpO₂ recorded during the examination was higher in the intervention group (94.2 ± 2.1% vs. 90.5 ± 3.8%, P < 0.05). The incidence of severe hypoxemia (SpO₂ < 85%) was also lower in the intervention group (P < 0.05). In addition, the incidence of hypoxemic events requiring clinical intervention was significantly lower in the intervention group (P < 0.05; Table 2).

Table 2.

Comparison of hypoxemia-related parameters between the two groups

Parameters	Intervention group (n = 67)	Control group (n = 67)	t/χ²	P
Incidence of hypoxemia, n(%)	8 (11.9)	25 (37.3)	χ²=12.381	< 0.001
Lowest SpO₂ (%)	94.2 ± 2.1	90.5 ± 3.8	t = 7.245	< 0.001
Incidence of severe hypoxemia, n(%)	1 (1.5)	9 (13.4)	χ²=6.782	0.009
Hypoxemic events requiring intervention, n(%)	5 (7.5)	18 (26.9)	χ²=9.120	0.003

Analysis of independent risk factors for hypoxemia

Patients who had received personalized positioning management were independently associated with lower risk of hypoxemia during sedated endoscopic examination (OR = 0.22, 95% CI: 0.09–0.54, P = 0.001). A higher ASA-PS (Class III) was independently associated with an increased risk of hypoxemia (OR = 2.85, 95% CI: 1.21–6.69, P = 0.016). Age, sex, BMI, comorbid COPD, and smoking history were not independently associated with the occurrence of hypoxemia (all P > 0.05; Table 3).

Table 3.

Logistic regression analysis of factors associated with hypoxemia

Variables	β	S.E	Wald χ²	P	OR [95%CI]
Group allocation (Control group)	-1.53	0.47	10.58	0.001	0.22[0.09–0.54]
Age (years)	0.04	0.04	1.00	0.317	1.04[0.96–1.13]
Sex (Male)	-0.15	0.45	0.11	0.740	0.86[0.36–2.08]
BMI (kg/m²)	0.06	0.07	0.74	0.390	1.06[0.93–1.22]
ASA-PS classification (Class Ⅲ)	1.05	0.44	5.70	0.016	2.85[1.21–6.69]
Comorbid COPD (yes)	0.62	0.55	1.27	0.260	1.86[0.63–5.45]

To further explore whether the difference in propofol dosage between groups influenced the observed outcomes, an additional multivariable logistic regression analysis was performed including propofol dosage and procedure time as covariates, in addition to all variables in the primary model (group allocation, age, gender, BMI, ASA-PS, comorbid COPD, and smoking history). After adjustment for these factors, personalized positioning management remained significantly associated with a lower risk of hypoxemia (OR = 0.142, 95% CI: 0.042–0.416, P < 0.001). Propofol dosage (P = 0.127) and procedure time (P = 0.537) were not independently associated with hypoxemia in this model. The full regression results are presented in Supplementary Table S1.

Patient comfort and procedural efficiency

VAS results demonstrated that postoperative comfort scores were significantly higher in the intervention group than in the control group (P < 0.05). The mean endoscopic procedure time in the intervention group was 8.5 ± 2.1 min, which was significantly shorter than that in the control group (10.2 ± 3.0 min; t = 4.012, P < 0.05, Table 4).

Table 4.

Comparison of intraoperative interventions and procedural efficiency between the two groups

Parameters	Intervention group (n = 67)	Control group (n = 67)	t	P
Endoscopic procedure time (min)	8.5 ± 2.1	10.2 ± 3.0	t = 4.012	< 0.001
Postoperative comfort scores	8.5 ± 1.2	7.1 ± 1.6	t = 5.674	< 0.001

Adverse events and safety analysis

The magnitude of heart rate fluctuation and systolic blood pressure variation during the procedure was significantly smaller in the intervention group (both P < 0.001). The overall incidence of intraoperative adverse events was lower in the intervention group (P < 0.05). Notably, no episodes of transient bradycardia occurred in the intervention group. All adverse events were promptly managed and resolved without progression to serious complications. The total propofol dosage in the intervention group was 115.3 ± 25.6 mg, which was significantly lower than that in the control group (128.7 ± 30.1 mg; t = 2.782, P = 0.006). The proportion of patients requiring airway intervention was 4.5% (3 cases) in the intervention group, lower than the 16.4% (11 cases) observed in the control group (χ² = 5.547, P = 0.019, Table 5).

Table 5.

Comparison of patient comfort and safety outcomes between the two groups

Parameters	Intervention group (n = 67)	Control group (n = 67)	t/χ²	P
Heart rate fluctuation (times/min)	12.3 ± 3.5	18.6 ± 4.8	t = 8.912	< 0.001
Systolic blood pressure variation (mmHg)	15.2 ± 4.1	22.8 ± 5.7	t = 8.634	< 0.001
Overall incidence of intraoperative adverse events, n(%)	5 (7.5)	16 (23.9)	χ²=7.264	0.007
Body movement response	3 (4.5)	8 (11.9)
Nausea and vomiting	2 (3.0)	5 (7.5)
Transient bradycardia	0 (0.0)	3 (4.5)
Total propofol dosage (mg)	115.3 ± 25.6	128.7 ± 30.1	t = 2.782	0.006
Proportion of patients requiring airway intervention, n(%)	3 (4.5)	11 (16.4)	χ²=5.547	0.019

Discussion

In this retrospective analysis, personalized positioning management was associated with a lower incidence and severity of hypoxemia during sedated endoscopic examination in elderly patients with EC. The observed incidence of hypoxemia was lower in the intervention group than in the control group, the lowest SpO₂ values were higher, and the incidence of hypoxemic events requiring intervention was fewer. Additionally, the personalized positioning group appeared to have shorter examination times, lower cumulative propofol doses, and higher postoperative comfort scores. The overall incidence of adverse events was also lower in the intervention group, with no episodes of transient bradycardia recorded.

The core mechanism by which personalized positioning management prevents hypoxemia lies in its optimization of respiratory mechanics. Elderly patients with EC often exhibit age-related decline in chest wall compliance and respiratory muscle strength, which may be further exacerbated by tumor-induced airway compression or poor nutritional status., leading to decreased functional residual capacity. Under sedation, these factors predispose patients to alveolar collapse and ventilation–perfusion mismatch. In this study, the individualized positioning protocol maintained a naturally aligned airway axis, minimized body torsion, and reduced diaphragmatic compression by abdominal contents, thereby improving ventilatory efficiency. This optimization may effectively reduce hypoventilation-related hypercapnia and acidosis. Moreover, improved positioning reduces compensatory respiratory muscle workload and cardiovascular stress responses, while a stable respiratory sensation and reduced physical restraint can significantly enhance patients’ psychological sense of safety [17].

The effectiveness of positional adjustment in improving ventilation has been supported by multiple studies. Kobayashi et al. [18] demonstrated that head-elevated positioning significantly reduces passive upper airway collapsibility during propofol anesthesia, which is consistent with the rationale of optimizing airway alignment through head elevation in the present study. In addition, Hedenstierna et al. [19] reported that elevating the upper body by 25°–30° in obese patients effectively improves respiratory mechanics and oxygenation, corroborating the significantly higher lowest SpO₂ observed in the intervention group in this study. Importantly, our findings indicate that personalized positioning management not only significantly reduces the incidence of hypoxemia but also markedly attenuates its severity. This effect is directly related to the physiological mechanism by which positional optimization enhances chest wall mobility and lung volumes, thereby increasing oxygen reserves [20]. Kobayashi et al. [18] also reported a close association between patient positioning, atelectasis, and gas exchange during anesthesia. In the present study, the incidence of severe hypoxemia (SpO₂ < 85%) in the intervention group was only 1.5%. This outcome may be attributed, on the one hand, to a reduced risk of upper airway collapse during sedation, and on the other hand, to improved oxygen reserves resulting from enhanced chest wall mobility. Furthermore, the significantly shorter examination time and reduced propofol dosage observed in the intervention group suggest that positioning management may influence drug distribution and metabolism. Improved ventilation efficiency and more stable oxygenation likely reduced the need for frequent adjustments in anesthetic depth due to hypoxemic events, while specific body positions may also affect drug distribution and clearance [21, 22]. Wani et al. [14] similarly reported that patient positioning is associated with sedative requirements and complication risk, which is consistent with the smaller fluctuations in heart rate and blood pressure and the lower overall incidence of adverse events observed in the intervention group. Mezidi et al. [23] further found that lung-protective strategies, including positional adjustment, can improve respiratory function and reduce stress responses in obese patients, aligning with the more stable vital signs observed in this study. In addition, Heller et al. [24] demonstrated that appropriate positional support can significantly reduce patient anxiety during procedures.

This retrospective analysis indicates that personalized positioning management may be a feasible and practical strategy to improve peri-procedural outcomes in elderly patients undergoing sedated endoscopic examination. By potentially reducing the risk of hypoxemia and adverse events, while enhancing procedural efficiency and patient comfort, this approach could help improve patient tolerance and adherence to endoscopic procedures. Its simplicity, lack of need for expensive equipment, and minimal additional training requirements make it a clinically implementable strategy.

Although this study did not quantitatively assess endoscopic examination quality (e.g., biopsy success rate or image clarity), all procedures in both groups were completed successfully without premature termination. No patient required procedure abandonment due to positioning-related difficulties. The shorter procedure time and lower propofol dosage in the intervention group did not compromise the completion of planned examinations. Nevertheless, the lack of standardized quality metrics is a limitation of this retrospective study, and future prospective studies should include formal assessment of examination quality.

Additionally, although all patients received supplemental oxygen at 5 L/min according to the institutional protocol, oxygen flow rate was not individually adjusted or recorded for each patient due to the retrospective design. This may have influenced the observed incidence and severity of hypoxemia. Tumor stage data were not available in this retrospective cohort, which we acknowledge as a limitation. Tumor stage may influence procedural risk through mechanisms such as airway compression, aspiration, or reflux-related respiratory events during sedation. However, several factors partially mitigate this concern. First, in accordance with our institutional practice and national sedation guidelines, patients with known severe airway compression or mediastinal invasion documented by prior imaging were generally not referred for sedated endoscopy; such patients were preferentially examined under topical anesthesia or with anesthesia standby. This practice likely reduced the inclusion of patients with the most severe tumor-related airway compromise in both groups. Second, all baseline characteristics, including age, ASA classification, and comorbidities, were well balanced between the two groups, suggesting that unmeasured tumor-related factors were unlikely to have differentially affected the groups. Nonetheless, the absence of detailed tumor staging precludes definitive adjustment for this potential confounder, and future prospective studies should incorporate staging information to further clarify its relationship with hypoxemia risk during sedated endoscopy. However, these findings are subject to selection bias due to the single-center nature, and may have limited generalizability. Documentation quality, operator experience, and variations in procedural execution could influence the observed associations. Moreover, electrical impedance tomography was not available. Prospective studies, preferably randomized controlled trials, are needed to further clarify the potential benefits and mechanisms of personalized positioning management and to establish evidence-based guidelines.

Whether these findings can be generalized to younger patients or those with other gastrointestinal malignancies remains unknown. Younger patients may have better respiratory reserve and lower baseline hypoxemia risk, potentially reducing the magnitude of benefit from personalized positioning. Further studies in diverse populations are needed to validate the external validity of our findings.

In conclusion, this retrospective analysis observed that personalized positioning management was associated with improved oxygenation, shorter procedural times, reduced sedative requirements, and higher patient comfort scores in elderly patients undergoing sedated endoscopic examination for EC. These observations suggest that optimizing patient positioning may provide a feasible strategy to enhance procedural safety, efficiency, and patient experience, although causal relationships cannot be established based on the current study design.

Authors' contributions

Yuan Chen, Jia Zhu designed the study, reviewed the manuscript, conducted the analysis, drafted the manuscript, contributed to the data visualization. and approved the final version.

Funding

None.

Data availability

The data presented in this study are available on request from the corresponding author.

Declarations

Ethical approval and consent to participate

The study protocol was approved by the Institutional Ethics Committee of Sir Run Run Shaw Hospital, Zhejiang University School of Medicine (No. 2025026) and conducted following the Declaration of Helsinki. Due to the study’s retrospective design, the ethics committee of Sir Run Run Shaw Hospital, Zhejiang University School of Medicine granted a waiver for obtaining written informed consent.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.