Linear Mixed-Effects Models in Medical Research

Schober, Patrick; Vetter, Thomas R.

doi:10.1213/ane.0000000000005541

Cited by 14 publications

(18 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Although this element may be innovative compared with other studies, we felt it was a more patient-centred design in terms of the patients’ daily routine. Third, with longitudinal data, such as postoperative pain intensity, the GLMM is the most appropriate choice, despite the nonnormal distribution of our data 30 . Unfortunately, no nonparametric alternative exists for a two-factor mixed model.…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Comparison between supra-inguinal fascia iliaca and pericapsular nerve group blocks on postoperative pain and functional recovery after total hip arthroplasty

Carella

Beck

Piette

et al. 2023

European Journal of Anaesthesiology

View full text Add to dashboard Cite

BACKGROUND Pain after a posterolateral approach for total hip arthroplasty (THA) may affect early functional recovery. Supra-inguinal fascia iliaca (SFIB) and pericapsular nerve group (PENG) blocks have been proposed as promising analgesia techniques. OBJECTIVES This trial was conducted to compare a PENG with a SFIB for controlling postoperative pain and for providing functional recovery. DESIGN Noninferiority monocentric randomised controlled study. SETTING One hundred and two patients scheduled for a total hip arthroplasty via the posterolateral approach under spinal anaesthesia were prospectively allocated to two groups. Data acquisition occurred between October 2021 and July 2022 at the University Hospital of Liege. PATIENTS One hundred and two patients completed the trial. INTERVENTIONS Group SFIB received supra-inguinal fascia iliaca block (SFIB) (40 ml ropivacaine 0.375%), whereas group PENG received a PENG block (20 ml ropivacaine 0.75%). MAIN OUTCOME MEASURES Rest and mobilisation pain on a 0 to 10 numeric rating scale at fixed time points: 1 and 6 h after surgery, on day-1 and day-2 at 8 a.m.,1 p.m. and 6 p.m. On day-1 and day-2, evolution of quality-of-recovery-15 score was assessed, and timed-up-and-go, 2 and 6 min-walking tests. The noninferiority margin was set as 1 numeric rating scale point 6 h after surgery. RESULTS Six hours after surgery, pain scores in group PENG were noninferior to those of group SFIB, with a difference between medians at 0 (95% CI −0.93 to 0.93). There were no significant differences between the groups regarding rest and dynamic pain trajectories during the first 48 postoperative hours, with no significant effects of group (rest P = 0.800; dynamic P = 0.708) or interaction between group and time (rest P = 0.803; dynamic P = 0.187). Similarly, no significant differences were observed regarding motor and functional recovery as assessed by timed-up-and-go (P = 0.197), 2 min (P = 0.364), and 6 min walking (P = 0.347) tests and quality-of-recovery-15 (P = 0.417) score. CONCLUSION Following a total hip arthroplasty via the posterolateral approach, a PENG block is noninferior to SFIB regarding postoperative pain control 6 h after surgery, and functional recovery. TRIAL REGISTRATION European Clinical Trial Register under EudraCT-number 2020-005126-28 (https://www.clinicaltrialsregister.eu/ctr-search/trial/2020-005126-28/BE).

show abstract

Section: Discussionmentioning

confidence: 99%

“…Third, with longitudinal data, such as postoperative pain intensity, the GLMM is the most appropriate choice, despite the nonnormal distribution of our data. 30 Unfortunately, no nonparametric alternative exists for a two-factor mixed model. Finally, no determination of isometric muscle strength was assessed after the peripheral blocks.…”

Section: Discussionmentioning

confidence: 99%

Comparison between supra-inguinal fascia iliaca and pericapsular nerve group blocks on postoperative pain and functional recovery after total hip arthroplasty

Carella

Beck

Piette

et al. 2023

European Journal of Anaesthesiology

View full text Add to dashboard Cite

show abstract

“…Examples of studies that require mixed-effects models are time series, longitudinal studies with multiple measurements taken over time on the same subject, and studies that involve participants from different sites or that are evaluated by different physicians (clustered measurements). In such cases, the effects of participants and centers would be considered a random effect, and they would be added to the model to take into account correlations between measurements 2 that traditional linear or logistic regression cannot address.…”

Section: Characteristics and Usefulness Of Mixed-effects Modelsmentioning

confidence: 99%

Mixed-effects model: a useful statistical tool for longitudinal and cluster studies

2023

View full text Add to dashboard Cite

“…The framework of LME models also performs ‘shrinkage’ for estimating model parameters; that is, individual estimates obtained from LME models are shrunk towards a grand mean of the population level estimate compared to fitting separate linear models to each subject’s data ( Bell et al, 2019 ). ME models have a long history of use in health and medicine since these models treat each patient not only as a member of a population but as an individual with unique characteristics ( Gelman et al, 2012 ; Barr et al, 2013 ; Baldwin et al, 2014 ; Wang et al, 2019 ; Schober and Vetter, 2021 ). ME models thus allow estimating model parameters that describe between- and within-subject variability of individual responses.…”

Section: Introductionmentioning

confidence: 99%

Pune GSH supplementation study: Analyzing longitudinal changes in type 2 diabetic patients using linear mixed-effects models

Madathil

Ghaskadbi²,

Kalamkar³

et al. 2023

Front. Pharmacol.

View full text Add to dashboard Cite

Oral GSH supplementation along with antidiabetic treatment was shown to restore the body stores of GSH significantly and reduce oxidative DNA damage (8-OHdG) in Indian Type 2 diabetic (T2D) patients over 6 months in our recent clinical study. Post hoc analysis of the data also suggested that elder patients benefit from improved HbA1c and fasting insulin. We modeled longitudinal changes in diabetic individuals using a linear mixed-effects (LME) framework and obtained i) the distribution of individual trajectories with and without GSH supplementation and ii) the overall rates of changes in the different study arms. Serial changes in elder and younger diabetic individuals were also modeled independently to examine differences in their progression. The average linear trajectories obtained from the model explain how biochemical parameters in T2D patients progress over 6 months on GSH supplementation. Model estimates show improvements in erythrocytic GSH of 108 µM per month and a reduction in 8-OHdG at a rate of 18.5 ng/μg DNA per month in T2D patients. GSH replenishes faster in younger people than in the elder. 8-OHdG reduced more rapidly in the elder (24 ng/μg DNA per month) than in younger (12 ng/μg DNA per month) individuals. Interestingly, elder individuals show a substantial reduction in HbA1c (0.1% per month) and increased fasting insulin (0.6 µU/mL per month). Changes in GSH correlate strongly with changes in HbA1c, 8-OHdG, and fasting insulin in the elder cohort. The model estimates strongly suggest it improves the rate of replenishment in erythrocytic GSH stores and reduces oxidative DNA damage. Elder and younger T2D patients respond differently to GSH supplementation: It improves the rate of reduction in HbA1c and increases fasting insulin in elder patients. These model forecasts have clinical implications that aid in personalizing treatment targets for using oral GSH as adjuvant therapy in diabetes.

show abstract

Linear Mixed-Effects Models in Medical Research

Cited by 14 publications

References 6 publications

Comparison between supra-inguinal fascia iliaca and pericapsular nerve group blocks on postoperative pain and functional recovery after total hip arthroplasty

Comparison between supra-inguinal fascia iliaca and pericapsular nerve group blocks on postoperative pain and functional recovery after total hip arthroplasty

Mixed-effects model: a useful statistical tool for longitudinal and cluster studies

Pune GSH supplementation study: Analyzing longitudinal changes in type 2 diabetic patients using linear mixed-effects models

Contact Info

Product

Resources

About