Introduction
With advancements in medical models, day surgery has become the mainstream approach for elective procedures worldwide, offering key benefits such as efficient allocation of medical resources, accelerated postoperative recovery and improved patient satisfaction.1 Data from developed countries indicate that day surgeries account for more than 60%–70% of all elective procedures.2–4 While this transformative model enhances medical efficiency, it also introduces new challenges in perioperative management. Despite continuous progress in minimally invasive techniques and precise anaesthesia management, postoperative pain remains a significant factor limiting the quality of day surgery. Over the past 30 years, the incidence of severe pain on the first postoperative day has consistently remained at approximately 20%.5 A study on day surgery in Germany reported that the average incidence of severe pain in the first 3 days postoperatively is 22.5%6 and, by the seventh day, postoperative pain persists in 18% of patients.7 Uncontrolled postoperative pain significantly reduces patient satisfaction; disrupts daily life, work and recovery; and may contribute to psychological conditions such as anxiety, depression and sleep disturbances.8–10 Additionally, inadequate pain management over an extended period can increase the risk of developing chronic pain11 and, in some cases, necessitate hospital readmission for further treatment.12
With the rapid expansion of day surgery, postoperative pain management has become a critical factor influencing patient recovery and satisfaction. One possible reason is that day surgery encompasses a wide range of surgical procedures, resulting in significant variations in postoperative pain levels.6 Additionally, substantial physiological and psychological differences among patients necessitate a highly individualised approach to pain management.13 While day surgery reduces hospital stays, it also shifts the crucial recovery period (3–7 days) for pain monitoring and intervention to the home setting. After discharge, patients often struggle to communicate with medical staff in a timely manner and may have difficulty receiving adequate support. Medical staff typically provide postoperative pain management guidance based on a patient’s pain perception at discharge; however, studies suggest that this approach is not always effective.14 Research indicates that recovery time following day surgery can range from a few days to several months.15 Current studies on postoperative pain in day surgery often focus on pain at a single time point, overlooking its dynamic progression and providing only symptomatic treatment. During recovery, pain trajectories vary—some patients experience gradual relief, while others develop chronic pain.11 Traditional research treats pain as merely one of the indicators of recovery, failing to explore the potential relationship between pain and overall recovery outcomes.
Therefore, we hypothesise that distinct postoperative pain trajectories are significantly associated with postoperative recovery outcomes. To address this, we have designed a prospective cohort study to assess patients’ pain at multiple time points within the first month after surgery. Using trajectory analysis, we aim to classify and identify distinct pain trajectories. Additionally, we will examine the relationship between different pain trajectory categories and recovery outcomes, including patients’ daily activities, work and emotional well-being. By analysing the dynamic changes in pain following day surgery, this study seeks to shift postoperative management from a ‘symptom management’ approach to a ‘comprehensive intervention’ model. This trajectory-oriented framework has the potential to transform both clinical practice and research, offering a more personalised and proactive approach to postoperative care.
Research objectives
-
To identify and classify postoperative pain trajectories over 1 month in day surgery patients and dynamically observe pain progression during the recovery period.
-
To explore the relationship between different pain trajectory categories and postoperative recovery, with a focus on the impact of pain on daily life, functional outcomes and emotional well-being.
-
To investigate potential risk factors associated with high-risk postoperative pain trajectories, providing a scientific basis for the development of personalised and precise pain management strategies.
Method
Study design
This study is designed as a prospective cohort study. All participants will undergo multiple pain assessments within the first month postoperatively to capture the dynamic changes in pain during recovery from day surgery.
Patient characteristics will be collected across multiple dimensions, including personal information, preoperative physical and mental status, surgical details, medication use and adverse reactions. The study protocol has been designed and reported in accordance with the Standard Protocol Items: Recommendations for Interventional Trials guidelines to ensure transparency and methodological rigor.16
Inclusion criteria
Exclusion criteria
-
Patients with communication barriers that prevent effective interaction, including hearing impairment, visual impairment, cognitive impairment or emotional disorders.
-
Patients with pre-existing pain conditions unrelated to the day surgery (eg, pain associated with malignant tumours or coexisting conditions such as arthritis).
Withdrawal criterion
Data collection
On patient admission, the investigator will visit the bedside to obtain written informed consent. Basic information, including demographic details, medical history, a comprehensive pain assessment and evaluations of emotional well-being and sleep quality, will be collected through face-to-face interviews. Postoperatively, anaesthesia and surgical details—such as the type of procedure, duration and analgesic methods—will be recorded using the electronic medical record system. Follow-up assessments will be conducted via telephone on postoperative days one, three and seven and at 1 month, during which pain levels and recovery status will be evaluated. The specific follow-up time points are illustrated in figure 1.
During the preoperative follow-up, we will record patients’ demographic details, including sex, age, height, weight, education level, marital status, smoking history and alcohol consumption history. Additionally, we will document a comprehensive medical history, including any chronic conditions, medications, allergies and previous surgeries. To assess patients’ emotional well-being, we will use the Hospital Anxiety and Depression Scale (HADS), a validated tool designed to evaluate symptoms of anxiety and depression.17 The HADS consists of 14 items, divided into two subscales: one measuring anxiety and the other measuring depression. Known for its simplicity and efficiency, it is widely used for perioperative emotional assessments. A total score greater than eight on either the anxiety or depression subscale is considered positive, indicating the presence of anxiety or depressive symptoms. Sleep quality will be evaluated using the Pittsburgh Sleep Quality Index (PSQI), a widely used and validated questionnaire assessing sleep patterns over the past month.18 The PSQI consists of 19 self-rated items covering multiple aspects of sleep, including sleep quality, latency, duration, efficiency, disturbances and the use of sleep medications. A total PSQI score of ≥5 is generally considered indicative of sleep disturbances, reflecting impaired sleep quality.
Pain assessment
Before surgery, we will ask patients whether they experience chronic or acute pain. If preoperative pain is present, they will be required to complete the Brief Pain Inventory (BPI) scale. The BPI will also be used at all four postoperative follow-up points to assess pain levels. The BPI consists of two primary components17:
-
Pain severity: this section evaluates four key aspects of pain: worst pain, least pain, average pain and current pain intensity. Each is rated on a 0–10 scale, with higher scores indicating greater pain severity.
-
Impact of pain on quality of life: this section assesses how pain affects daily activities, sleep, work and overall quality of life through seven items, each scored from 0–10, where higher scores indicate a greater impact.
In this study, the postoperative pain score will be derived from the ‘average pain over the past 24 hours’ item, while the breakthrough pain score will be based on the ‘worst pain over the past 24 hours’ item. Additionally, the use of analgesic medications will be systematically recorded both preoperatively and postoperatively. Data will include the name of the medication, dosage, duration of administration and the patient’s self-reported assessment of analgesic efficacy. Pain relief effectiveness will be rated on a 0–10 scale, with higher scores indicating greater pain relief. To capture pain trajectories, assessments will be conducted on postoperative days 1, 3, 7 and 30.
Recovery assessment
We will evaluate patient recovery within the first month after surgery. The Recovery Quality of Life Scale-15 (QoR-15)19 will be used to assess postoperative recovery at the 1-month follow-up. This validated tool measures multiple dimensions of postoperative rehabilitation, including functional recovery, pain control, emotional well-being and overall health perception. At each postoperative follow-up, patients will also be asked to assess their own recovery status using a Subjective Rehabilitation Score. Patients will rate their physical recovery on a scale from 0–10, with higher scores indicating better recovery. Additionally, we will inquire about any postoperative complications experienced, including nausea and vomiting, urinary retention, cognitive impairment, poor wound healing, bleeding, abdominal distension, difficulty in defecation or passing gas and other surgery-related or anaesthesia-related complications.
Satisfaction score
Patients will rate their overall satisfaction with the treatment process and the service provided by healthcare professionals on a scale from 0 to 10, with higher scores indicating greater satisfaction.
Anaesthesia and surgical information
Anaesthesia and surgical details will be extracted from the patient’s electronic medical record, including American Society of Anesthesiologists (ASA) classification, anaesthesia modality, duration of anaesthesia, postoperative pain management strategy, type of surgery and estimated blood loss.
Study outcome measures
The primary outcome measure of this study is the recovery status at 1 month postoperatively, assessed using the QoR-15. The secondary outcome measures are the patient’s subjective recovery rating, satisfaction score and pain interference with daily life score. All questionnaires will be collected through an electronic information system, with specific scales and time points detailed in table 1.
Participant timeline
Quality control
The research team has undergone comprehensive standardised training on the use of assessment scales to ensure consistency in questionnaire administration. A dedicated quality control committee has been established to oversee the study, provide guidance and conduct random audits to maintain data integrity and a high follow-up rate within the study cohort. All data were deidentified immediately after collection by removing direct identifiers (eg, names, inpatient ID) and were stored on a designated encrypted computer, accessible only to the research team for data analysis. This study follows protocol V.1.0, finalised on 16 January 2025, with any future amendments documented through version updates and corresponding dates.
Data analysis
This study aims to identify postoperative pain trajectory categories within 1 month in day surgery patients and explore the relationship between different trajectory categories and rehabilitation. The analysis plan involves reporting the baseline characteristics of the study population, including demographic characteristics (eg, age, sex, body mass index) and clinical characteristics (eg, type of surgery, medical history). Continuous variables will be presented as mean (SD) or median (IQR), while categorical variables will be presented as frequency (percentage).
The average scores of postoperative pain and breakthrough pain will be calculated at four time points (day 1, day 3, day 7 and 1 month). The incidence of moderate to severe pain and moderate to severe breakthrough pain will also be reported. Patients will be stratified by surgery type, sex and postoperative analgesia method, and χ2 tests will be used to compare differences in pain incidence between subgroups. Using the BPI, we will calculate the average pain-related life interference scores across various dimensions, as well as the total average score across all seven dimensions to assess the overall impact of pain on quality of life.20 Descriptive statistics will be performed for postoperative QoR-15 scores, satisfaction scores and subjective recovery scores at the four follow-up time points. Patients will be stratified based on the presence of moderate to severe postoperative pain, and comparisons of QoR-15 scores between groups will be conducted using an independent samples t-test or Mann-Whitney U test, depending on whether the data follow a normal distribution. All collected data will be used for statistical analyses. Missing values will be imputed using the last observation carried forward method. Additionally, a sensitivity analysis will be conducted using a complete-case approach, excluding patients with missing data.
Postoperative pain trajectory analysis
To explore the dynamic changes in postoperative pain and their relationship with recovery outcomes, latent class growth analysis (LCGA) will be used. LCGA is a statistical model designed for longitudinal data,21 allowing for the identification of pain trajectory categories at different time points and the classification of individuals into distinct latent classes. LCGA will be applied to categorise pain scores at postoperative day 1, day 3, day 7 and 1 month. The number of latent classes will be determined using maximum likelihood estimation and evaluated based on the Bayesian information criterion (BIC for model fit, with the model yielding the lowest BIC value selected. To ensure statistical validity, all latent classes will include a group representation of ≥5%.
To compare differences in QoR-15 scores, satisfaction scores and pain interference scores (from the BPI questionnaire) between different pain trajectory classes at 1 month post surgery, one-way analysis of variance (ANOVA) will be conducted. Prior to performing ANOVA, tests for normality and homogeneity of variance will be conducted. If the data do not meet the assumptions of ANOVA, the Kruskal-Wallis test will be used as an alternative.
To further examine the relationship between pain trajectory classes and postoperative QoR-15 and satisfaction scores, we will construct a multiple linear regression model, adjusting for potential confounders such as age, sex, ASA grade, surgery duration and preoperative health status. In this regression analysis, latent classes will be treated as independent variables, while QoR-15 and satisfaction scores will serve as dependent variables.
Identification of risk factors
If trajectory and correlation analyses identify high-risk pain trajectory classes associated with poor recovery or low satisfaction, we will use generalised linear mixed models (GLMMs) to identify potential risk factors for these groups. GLMM is a flexible statistical model capable of handling both fixed effects (eg, patient characteristics, surgery type) and random effects (eg, individual differences, time variation), making it particularly well-suited for the analysis of repeated measurement data.22
The final model will output ORs and 95% CIs for each risk factor to assess its independent impact on postoperative pain. To prevent overfitting and improve variable selection, we will apply least absolute shrinkage and selection operator (LASSO) regression23 for feature selection. LASSO regression will use 10-fold cross-validation to identify the most predictive features, further optimising model performance. The data analysis process will iteratively refine the model to enhance its fit and predictive accuracy. Model performance will be validated using receiver operating characteristic curves and the area under the curve. Additionally, to further improve predictive accuracy, random forests or support vector machines may be considered as alternative modelling approaches.
All statistical analyses will be conducted using R software (V.4.2.1). Two-sided tests will be applied, with a significance level set at α=0.05.
Sample size
This prospective cohort study aims to enrol 1000 patients undergoing day surgery at Sichuan Provincial People’s Hospital over a 6-month period. In LCGA, a sample size of 1000 is generally sufficient to ensure adequate statistical power and model stability, particularly when class enumeration is aided by the BIC and bootstrap likelihood ratio test across varying model complexities and class distributions.24
Ethics and dissemination
This study has been approved by the ethics committee of Sichuan Provincial People’s Hospital (ethical review No. 749, 2024, contact email address: kejiaochu2006@126.com). Written informed consent will be obtained from all participants before their inclusion in the study. Participants will be fully informed about the study’s purpose, procedures and potential risks and will have the opportunity to ask questions before providing consent. All data will be anonymised, and participants’ personal information will be kept confidential in compliance with data protection regulations.
The findings of this study will be disseminated through peer-reviewed publications in academic journals and presentations at relevant conferences. Results will also be shared with healthcare professionals and policymakers to inform clinical practice. To ensure broad accessibility, an open-access publication model will be used. De-identified data that do not involve patient privacy will be available for sharing. To request access, interested researchers may contact the corresponding author via email and submit a formal written application. The anonymised dataset will be provided on approval by the ethics committee.
Patient and public involvement
The patients and the public had no role in the design, conduct, reporting or dissemination of the study.
Discussion
This prospective cohort study aims to explore postoperative pain trajectory categories in day surgery patients. Its innovation lies in a dynamic, multi-timepoint approach to pain observation. Unlike previous studies that focus solely on static pain assessments,12 25–27 this study considers pain occurrence and development as a whole. By conducting multiple pain assessments over a 1-month postoperative period, we employ LCGA to classify and identify different pain trajectory categories in day surgery patients. The LCGA model overcomes the limitations of traditional mean analysis by capturing the nonlinear characteristics of pain progression.28 A study of postoperative pain trajectories using the same statistical approach found that pain trajectories in the first 7 days after surgery were predominantly defined by patient factors and not surgical factors.29 Our study will further explore the effect of a longer period of pain trajectory on the rehabilitation of day surgery patients after discharge.
Moreover, in this study, pain trajectory categories are treated as independent variables, while recovery quality (QoR-15 score), patient satisfaction and other rehabilitation indicators serve as dependent variables to examine the long-term effects of different pain trajectories on postoperative recovery in day surgery patients. Previous research has primarily focused on pain at single time points, often overlooking its long-term progression. The key question in this study is whether different postoperative pain trajectory categories influence patients’ overall recovery. This approach fosters a shift in postoperative management from a ‘symptom response’ model to a ‘whole-process intervention’ strategy, introducing a ‘trajectory-guided’ framework for the comprehensive treatment of day surgery patients.
For potential high-risk subgroups of pain trajectories, this study will use GLMM to identify associated risk factors. GLMM can simultaneously account for fixed effects (eg, baseline patient characteristics, surgery type) and random effects (eg, individual differences, time variations), making it particularly suitable for repeated-measures data.30 In feature selection, this study innovatively incorporates anxiety, depression and sleep disturbances, recognising their potential role in the occurrence and progression of pain. Unlike conventional models that primarily focus on the presence of moderate to severe postoperative pain, this study uses high-risk pain trajectory class as an independent variable to identify risk factors. This approach may uncover distinct risk factors associated with specific dynamic pain classes, offering deeper insights into pain progression and management.
Limitations
This study has several limitations. First, it relies on self-reported data, which may be influenced by factors such as patient emotions, cultural background and comprehension ability, potentially affecting the accuracy of pain reports. To minimise this bias, we have enhanced investigator training and used simple, comprehensible assessment tools to ensure patients can accurately understand and express their pain experiences. Second, as a single-centre study, the findings may not fully represent a broader patient population. To enhance external validity and generalisability, future studies will aim to increase the sample size and collect data across multiple regions. Lastly, although more frequent assessments could provide a more precise delineation of pain progression, we selected only the essential time points (postoperative days 1, 3, 7 and 30) to balance methodological rigour with practical constraints. This minimised the risk of patient attrition associated with excessive follow-up burden.
This study provides valuable data and practical guidance for clinical pain management through an indepth analysis of postoperative pain trajectories in day surgery patients. Additionally, the findings may offer insights for optimising postoperative nursing care and rehabilitation programmes, contributing to improved patient recovery and satisfaction.
This post was originally published on https://bmjopen.bmj.com
