This single-center retrospective cohort evaluated whether the association between preoperative hemoglobin (Hb) and postoperative length of stay (LOS) after elective laparoscopic cholecystectomy (LC) varies by arterial oxygen saturation (SpO2) at high altitude. The authors note that Hb elevation is a characteristic high-altitude adaptation, but its clinical impact on surgical outcomes across differing oxygenation levels had not been previously studied. The investigators aimed to determine whether Hb effects on LOS reverse across predefined SpO2 strata and to characterize any such interaction.
The study included 612 adults who underwent elective LC at Qinghai Red Cross Hospital in Xining, China (altitude 2260 m) between 2018 and 2023. Key preoperative measures reported were Hb (range 82–233 g/L; mean 151.4 g/L; SD 20.7 g/L) and SpO2 (range 86%–99%; mean 94.6%; SD 2.3%). Demographic summary provided: 65.8% female and mean age 43.5 years (SD 11.9).
SpO2 was categorized into three strata for stratified analyses: low (<93%), mid (93%–95%), and high (≥96%). The primary analytic approach was multivariable linear regression of LOS with an interaction term between Hb and SpO2, adjusted for body mass index (BMI), age, sex, and season. The authors report the regression interaction coefficient (beta) and P value for statistical inference.
Across the 612 patients the investigators observed a statistically significant interaction between preoperative Hb and SpO2 on postoperative LOS (interaction beta = -0.0095; P = .009). This interaction indicates that the association between Hb and LOS differed depending on SpO2 level rather than being uniform across the sample.
Overall reported sample means were Hb 151.4 g/L and SpO2 94.6%, and the cohort was two-thirds female with a mean age of 43.5 years.
When effects were estimated within SpO2 strata the direction of association between Hb and LOS reversed. In the combined higher-SpO2 strata (SpO2 ≥93%), each 1 g/L increase in Hb was associated with a 0.003-day longer LOS (reported P = .079). Conversely, in the low-SpO2 stratum (SpO2 <93%), each 1 g/L increase in Hb was associated with a 0.006-day shorter LOS. The authors emphasize this reversal as the core observation and designate it the Plateau Hemoglobin Paradox.
A subgroup result reported for the mid-SpO2 stratum (n = 251) compared patients with very high Hb (≥180 g/L) to those with lower Hb: LOS was longer in the high-Hb group (1.88 days vs 1.62 days; P = .001) with an effect size d = 0.54.
The report states that five computational robustness analyses were performed and that these confirmed the Hb × SpO2 interaction. A leave-one-out analysis over the 612 observations reportedly yielded P < .05 in 100% of iterations, supporting stability of the interaction finding across individual-case removal tests. The manuscript provides these robustness checks as evidence that the observed interaction was not driven by single influential observations.
The authors describe this work as exploratory and note key limitations: it is a retrospective single-center study conducted at one altitude (2260 m) and therefore requires replication in independent cohorts. The achieved statistical power is reported as 0.754. The preprint status is explicitly noted: the article had not undergone peer review at the time of posting and should not be used to guide clinical practice.
Ethical approval was reported from the Ethics Committee of Qinghai Red Cross Hospital (Approval No. LW-2026-73). The authors declared no competing interests. Data availability was stated: data supporting the findings are available from the corresponding author on reasonable request.
In this cohort of patients undergoing elective LC at high altitude, the association between preoperative hemoglobin and postoperative length of stay varied by arterial oxygen saturation, with a statistically significant Hb × SpO2 interaction. The authors term the observed SpO2-dependent reversal the "Plateau Hemoglobin Paradox." They report that higher Hb related to shorter LOS when SpO2 was low (<93%), but to longer LOS in higher SpO2 strata, with a notable subgroup difference in the mid-SpO2 range for Hb ≥180 g/L.
The authors conclude that, if replicated, this pattern could inform perioperative risk stratification at high altitude. They emphasize the need for replication given the single-center design, the exploratory framing, the preprint status, and the reported statistical power below conventional thresholds for definitive inference. Details beyond those reported in the preprint (for example mechanistic data or external validation) were not provided by the source.