The clinical evidence-based practice questions were identified by integrating clinical experience with the PIPOST framework (8), which encompasses the following components: Population (P): Elderly patients with hip fractures; Intervention (I): Hypoxemia prevention and management protocol; Professionals (P): Orthopedic surgeons and nurses; Outcomes (O): Healthcare providers’ knowledge and awareness rates, postoperative SpO₂ fluctuations, and incidence of hypoxemia; Setting (S): Orthopedic center of a tertiary hospital in Ningxia; Type of Evidence (T): Clinical decision-making, guidelines, expert consensus, evidence summaries, and systematic reviews.

The evidence-based practice team comprised members with systematic training in evidence-based methodology, including: Three nursing researchers responsible for evidence translation, protocol development, data collection and analysis; One chief of trauma orthopedics and one resident physician who provided clinical guidance for evidence implementation and operational supervision; One nurse manager and two clinical nurse specialists from the trauma orthopedics ward overseeing protocol execution and quality control.

Guided by the 6S evidence pyramid model, we systematically searched for evidence on hypoxemia prevention and management in hip fracture patients across multiple databases, including The Cochrane Library, the National Institute for Health and Care Excellence (NICE) website, China National Knowledge Infrastructure (CNKI), Wanfang Database, and VIP Database. The search period extended from the inception of each database to March 2024. A total of seven publications were included, comprising four expert consensus documents and three clinical guidelines.

Three independent evaluators appraised the included expert consensus documents and clinical guidelines using the Professional Consensus Quality Assessment Tool and AGREE II instrument, respectively. Through this rigorous evaluation, six best-practice recommendations were synthesized.

Subsequently, a multidisciplinary panel meeting was convened, consisting of two trauma orthopedic surgeons, five nursing specialists, one anesthesiologist, and one pulmonologist. All panel members held associate senior titles or higher and possessed over 8 years of clinical experience. The panel critically evaluated the evidence across four key domains: feasibility, appropriateness, clinical significance, and effectiveness. During the discussion, specific operational modifications were proposed and deliberated for Recommendations 3–5, including: Intervention stratification: Development of tiered management protocols based on SpO₂ monitoring thresholds. Respiratory training optimization: Refinement of pulmonary rehabilitation exercises, specifying modalities, frequency (three times daily), and intensity (progressive increase to 30% of maximal inspiratory pressure) in accordance with Recommendation 5. The final six validated indicators were implemented into the clinical practice protocol (see Table 1).

This study used patients’ initial oxygen saturation measurement as the primary outcome indicator. The sample size was calculated based on the formula for comparing the means of two independent samples: n = (μ_α + μ_β)² × 2σ²/δ², set the test level of α = 0.05, β = 0.1. According to the standard normal distribution table, the corresponding values are μ_α = 1.96, and μ_β = 1.282. Taking the patients’ first oxygen saturation monitoring value as the main index, the investigated σ = 5.99, the proposed post-intervention increase in oxygen saturation values was 4. Substituting these values yielded a required sample size of n = 45 per group. Considering a 15% margin for potential attrition or calculation error, the final sample size was adjusted to 52 participants per group, resulting in a total sample size of 104.

A quasi-experimental, pre-post study design was utilized. Participants were assigned to groups based on the timing of their surgery: The pre-implementation (control) group consisted of consecutive eligible patients admitted between January and April 2024, who received routine standard care. The post-implementation (intervention) group consisted of consecutive eligible patients admitted between June and September 2024, who were managed according to the new evidence-based protocol. Inclusion criteria: ➀ age ≥ 65 years; ➁ Confirmed diagnosis of hip fracture; ➂ surgical intervention performed within 24 h of admission; ➃ patients who voluntarily agreed to participate in this study. Exclusion criteria: ➀ fractures caused by metastatic tumor; ➁ presence of rib fracture; ➂ intractable hypoxemia; ➃ severe cardiopulmonary diseases or cerebral hemorrhage; ➄ time of injury ≥ 2 weeks; ➅ operation duration ≥ 4 h; ➆ postoperative transfer to intensive care unit (ICU); ➇ American Society of Anesthesiologists (ASA) physical status classification grade IV or higher. This study was reviewed and approved by the Ethics Committee of Ningxia Hui Autonomous Region People’s Hospital (approval number: 2023-NZR-105). Informed consent was obtained from all participants prior to enrollment.

To control for potential confounding from variations in preoperative management, all patients in both the pre- and post-implementation groups were treated in accordance with the institution’s standardized clinical pathway for enhanced recovery after surgery (ERAS) in hip fracture patients.

Peripheral oxygen saturation (SpO₂) was measured using a Yuwell YX306 finger-clip pulse oximeter (Medical Device Registration Certificate No.: Su Zhun Zhun 20172071070). To minimize the impact of common interfering factors and ensure the acquisition of reliable data, a strict standardized protocol was implemented as follows: (1) Patient Preparation and State: Measurements were taken after the patient had been resting quietly in a supine or semi-Fowler’s position for at least 10 min in a non-oxygenated state. (2) Sensor Placement and Site Selection: The sensor was placed on the clean, warm, and unscarred distal phalanx of the index finger. Fingers with nail polish, artificial nails, or obvious deformities were avoided. The chosen site was checked for adequate local perfusion. This site was selected due to its widespread clinical use and practicality for repeated measurements. (3) Stabilization and Recording: The displayed SpO₂ value was recorded only after it had stabilized along with the waveform for a continuous period of 15–20 s. Transient fluctuations were ignored. (4) Environmental Controls: Strong ambient light sources (e.g., direct sunlight) were shielded from the sensor to prevent optical interference. This comprehensive quality control protocol was consistently applied by trained nursing staff to all measurements across both study groups, ensuring the comparability and reliability of the SpO₂ data.

Hypoxemia was defined according to the standard criteria (9): under room air conditions, SpO₂ < 90% was considered indicative of hypoxemia.

A test was developed to evaluate the knowledge of medical and nursing staff regarding the hypoxemia prevention and management program. To ensure its content validity and clinical relevance, the initial draft was reviewed by a multidisciplinary expert panel comprising two senior orthopedic surgeons, one pulmonologist, and three nursing experts. The panel evaluated the questions for clarity, appropriateness, and comprehensiveness in covering the key components of the protocol. Their feedback was used to refine the wording and content of the questionnaire. The test covers four key areas: concepts, classification, preventive measures, and emergency treatment. It consists of 20 questions, each worth 5 points, with a total score of 100 points. A higher score indicates a better mastery of the relevant knowledge. The questionnaire was pilot-tested with a group of 10 nurses not involved in the main study to identify any ambiguities in the questions or options, and minor revisions were made accordingly.

The most critical finding of our study is the dramatic reduction in hypoxemia incidence, from 13.46% (POD1) and 19.23% (POD2) in the control group to 1.92 and 0.00%, respectively, in the intervention group. This success can be attributed to the protocol’s proactive and dynamic design, which moves beyond conventional care that often discontinues monitoring after POD1 if SpO₂ > 90% (7). Our approach is consistent with recent international literature; for instance, a PRIME-AIR study protocol by Fernandez-Bustamante et al. concluded that bundled interventions, which combine monitoring, oxygen therapy, and respiratory training, are most effective at preventing postoperative pulmonary complications (14). Our protocol operationalizes this “bundle” concept into a precise, actionable clinical pathway. The significant “Group” effect and “Time × Group” interaction in our ANOVA further confirm that the intervention itself exerted a unique and powerful effect on stabilizing oxygenation over time. Although our study was not specifically designed for formal subgroup analyses by variables such as gender and age, several aspects of our data and analysis address this concern. First, as shown in Table 2, these key baseline characteristics were balanced between the pre-intervention and post-intervention groups. This balance reduces the likelihood that the observed significant differences in hypoxemia incidence and SpO₂ levels were influenced by these variables. Furthermore, repeated-measures ANOVA revealed a highly significant main effect of “Group” and a significant interaction between “Time × Group.” This indicates that the intervention itself exerts a robust and unique effect on the trajectory of oxygenation over time, an effect that is consistent across the study population. The primary objective of this study was to evaluate the overall efficacy of the bundled protocol; future research with larger sample sizes could help explore the moderating effects of specific patient or surgical factors on treatment outcomes.

Our data revealed that the highest incidence of hypoxemia in the control group occurred on POD2 (19.23%), a finding that underscores a commonly overlooked clinical window. This pattern aligns with the prospective study by Sun et al., which demonstrated that postoperative hypoxemia is not only common but also persistent, often extending beyond the immediate recovery period (9). Our protocol, by mandating a minimum of three days of dynamic monitoring, directly addresses this persistence and effectively prevents its occurrence. Regarding adverse events attributable to hypoxia in the pre-implementation group, no catastrophic events such as cardiac arrest or death were directly documented during the study period in either group. However, the significantly higher incidence of hypoxemia (SpO₂ < 90%) observed in the pre-implementation group (13.46% on POD1 and 19.23% on POD2) is itself a cause for grave concern. Extensive literature unequivocally links even transient postoperative hypoxemia to an increased risk of secondary complications, including myocardial injury, delirium, impaired wound healing, and prolonged hospitalization (1, 2, 4). The absence of immediate, life-threatening events in our cohort should not diminish the clinical significance of these findings; rather, it highlights that hypoxemia often operates as a ‘silent’ precursor to more serious morbidity. Therefore, the primary value of our evidence-based protocol lies in its proactive prevention of hypoxemia, thereby mitigating the risk cascade associated with it, rather than merely reacting to its most extreme consequences.

The combined use of oxygen therapy and structured respiratory training [e.g., incentive spirometry (15)] was a cornerstone of our protocol. This combination optimally engages respiratory muscles and helps prevent atelectasis. Our results strengthen the existing evidence (16) by demonstrating its efficacy within a comprehensive management strategy for a high-risk population. When compared to national studies, our pre-implementation hypoxemia rate was similar to the 15.5% reported by Liu et al. (3). However, our post-implementation results suggest that the comprehensive implementation strategy—including staff education, visual aids, and quality control—was crucial for achieving outcomes that surpass those of simpler interventions.