Reduction of self-perceived discomforts in critically ill patients in French intensive care units: study protocol for a cluster-randomized controlled trial

Measuring pain or discomfort during routine nursing care in lightly sedated mechanically ventilated intensive care patients: A prospective preliminary cohort study

BACKGROUND

Pain is routinely measured on mechanically ventilated ICU patients. However, the tools used are not designed to discriminate between pain and non-pain discomfort, a distinction with therapeutic implications.

OBJECTIVES

To evaluate whether clinical measurement tools can discern both pain and non-pain discomfort.

METHODS

A prospective observational cohort study was conducted in a General ICU at a tertiary Medical Center in Israel. The Behavior Pain Scale (BPS) and Visual Analog Scale (VAS) of Discomfort were simultaneously assessed by a researcher and bedside nurse on thirteen lightly sedated patients during 71 routine nursing interventions in lightly sedated, mechanically ventilated, adult patients. Patients were asked whether they were in pain due to these interventions.

RESULTS

Statistically significant increases from baseline during interventions were observed [median change: 1.00 (-1-5), 1.5(-4-8.5), p < 0.001] as measured by BPS and VAS Discomfort Scale, respectively. BPS scores ranged between 4 and 6 when the majority (53 %) of the patients replied that they had no pain but were interpreted by the clinicians as discomfort. Endotracheal suctioning caused the greatest increase in BPS and VAS, with no statistically significant differences in BPS and VAS Discomfort Scale scores whether patients reported or did not report pain. A BPS>6 had a higher sensitivity and specificity to reported pain (accuracy of 76 %) compared to a BPS of 4-6.

CONCLUSIONS

Standard assessments are sensitive to pain caused by routine nursing care interventions. However, this study presents evidence that among lightly sedated ICU patients, moderate BPS scores could also measure non-pain discomfort. ICU nurses should be aware that signs of unpleasantness measured by a pain scale could reflect non-pain discomfort.

Nurses' Attitudes, Practices, and Barriers to Assessing Symptoms of Discomfort in Mechanically Ventilated Patients: A Cross-Sectional Study

Introduction

Critically ill patients experience various stressful symptoms of discomfort, including dyspnea, pain, and sleep disruption. Notably, ventilated patients have difficulty self-reporting discomfort symptoms. Nurses need to assess discomfort symptoms to alleviate them, but limited research exists on discomfort symptom assessment and management in critically ill patients.

Objective

To identify the practices, attitudes, and barriers among nurses related to the assessment of discomfort symptoms in mechanically ventilated patients.

Methods

Using a cross-sectional, descriptive study design, a web-based survey was conducted between May and June 2022 with critical care nurses sampled through Japanese academic societies and social networking services. The survey contained questions relative to the above-stated objective. Descriptive statistical analysis was performed without sample size calculation because of the descriptive and exploratory nature of this study.

Results

There were 267 respondents to the questionnaire. The discomfort symptoms that nurses perceived as important to assess were pain (median 100 [interquartile range, IQR 90-100]), insomnia (99 [80-100]), and dyspnea (96.5 [75-100]). Most participants (89.8%) routinely assessed pain in mechanically ventilated patients using a scale; however, other discomfort symptoms were assessed by less than 40% (dyspnea [28.4%], fatigue [8.1%], thirst [13.1%], insomnia [37.3%], and anxiety [13.6%]). Two major barriers to assessing discomfort symptoms were lack of assessment culture within the intensive care unit and lack of knowledge of the relevant evaluation scales.

Conclusions

Nurses were aware of the importance of using scales to assess the discomfort symptoms experienced by mechanically ventilated patients. However, except for pain, most nurses did not routinely use scales to assess discomfort symptoms. Barriers to routine discomfort symptom assessment included the lack of an assessment culture and the lack of knowledge of the assessment scales. Clinicians should be educated regarding the existence of validated rating scales and develop additional rating scales utilizable for minor discomforts in mechanically ventilated patients.

An overview of optimal designs under a given budget in cluster randomized trials with a binary outcome

Cluster randomized trial design may raise financial concerns because the cost to recruit an additional cluster is much higher than to enroll an additional subject in subject-level randomized trials. Therefore, it is desirable to develop an optimal design. For local optimal designs, optimization means the minimum variance of the estimated treatment effect under the total budget. The local optimal design derived from the variance needs the input of an association parameter ρ in terms of a "working" correlation structure R ( ρ ) in the generalized estimating equation models. When the range of ρ instead of an exact value is available, the parameter space is defined as the range of ρ and the design space is defined as enrollment feasibility, for example, the number of clusters or cluster size. For any value ρ within the range, the optimal design and relative efficiency for each design in the design space is obtained. Then, for each design in the design space, the minimum relative efficiency within the parameter space is calculated. MaxiMin design is the optimal design that maximizes the minimum relative efficiency among all designs in the design space. Our contributions are threefold. First, for three common measures (risk difference, risk ratio, and odds ratio), we summarize all available local optimal designs and MaxiMin designs utilizing generalized estimating equation models when the group allocation proportion is predetermined for two-level and three-level parallel cluster randomized trials. We then propose the local optimal designs and MaxiMin designs using the same models when the group allocation proportion is undecided. Second, for partially nested designs, we develop the optimal designs for three common measures under the setting of equal number of subjects per cluster and exchangeable working correlation structure in the intervention group. Third, we create three new Statistical Analysis System (SAS) macros and update two existing SAS macros for all the optimal designs. We provide two examples to illustrate our methods.

Corporeal rehabilitation to manage acute stress in critically ill patients

Background

Intensive care unit (ICU) patients often endure discomfort and distress brought about by their medical environment and the subjective experience of their stay. Distress, pain, and loss of control are important predictors of future neuropsychiatric disorders. Depression, anxiety, and post-traumatic stress are common after discharge. We aimed at mitigating acute stress and discomfort via a novel intervention based on body image rehabilitation and rehabilitation of senses performed following a holistic approach guided by positive communication (corporeal rehabilitation care, CRC).

Results

We conducted a prospective observational study on 297 consecutively enrolled patients participating in at least one CRC session. Benefits of CRC were assessed on both subjective analogical scales of stress, pain, and well-being criteria, and objective clinical measures of dyspnea, respiratory rate, and systolic arterial pressure, just after CRC and long after (a median of 72 min later) to estimate its remote effect. Results showed that CRC had a positive effect on all overt measures of distress (acute stress, pain, discomfort) just after CRC and remotely. This beneficial effect was also observed on dyspnea and respiratory rate. Results also showed that best CRC responders had higher baseline values of stress and heart rate and lower baseline values of well-being score, indicating that the care targeted the population most at risk of developing psychological sequelae. Interestingly, a positive CRC response was associated with a better survival even after adjustment for physiologic severity, indicating a potential to identify patients prompt to better respond to other therapeutics and/or rehabilitation.

Conclusion

This study demonstrated the feasibility of an innovative holistic patient-centered care approach and its short-term positive effects on critical parameters that are considered risk factors for post-intensive care syndrome. Further studies are warranted to study long-term benefits for patients, and overall benefits for relatives as well as ICU staff.

Perception of discomfort by mechanical ventilation patients in the Intensive Care Unit: A qualitative study

BACKGROUND

Routine care in intensive care units (ICU) results in patient pain and discomfort. While pain is treated with analgesics, discomfort is generally not well characterised or addressed. Since many ICU patients communicate only non-verbally, practitioners often cannot discern between pain or discomfort when treating such patients, potentially leading to inappropriate analgesic administration. A first step in discriminating between pain and discomfort is understanding how patients perceive their discomfort.

OBJECTIVE

To describe mechanically ventilated ICU patients' perceptions of discomfort and how they differentiate discomfort from pain.

METHOD

A qualitative descriptive study using semi-structured interviews conducted with 13 patients in a Medical and General ICU who survived mechanical ventilation. Transcripts were analysed using content analysis.

FINDINGS

Two main discomfort themes were identified: unpleasant physical sensations and unpleasant psychological feelings. Each theme was further divided into subcategories. Most patients did not describe high levels of pain and did not associate physical discomfort with pain.

CONCLUSIONS

Discomfort, as described by patients, stems from both physical sensations and psychological feelings. Pain was less often described as a negative ICU experience, while other non-pain sources of discomfort were more likely to be recalled. Therefore, practitioners should not only focus on treating pain but also on treating overall comfort to improve the quality of the ICU experience and potentially decrease post-ICU psychological sequela.

Risk factors and events in the adult intensive care unit associated with pain as self-reported at the end of the intensive care unit stay

Background

The short-term and long-term consequences of the most frequent painful procedures performed in the ICU are unclear. This study aimed to identify the risk factors associated with pain-related discomfort perceived by critically ill patients during the whole ICU stay as self-reported by patients at the end of their ICU stay.

Methods

The study involved 34 ICUs. Adult patients who survived an ICU stay of 3 calendar days or more were eligible for inclusion. Discomforts, including the pain-related discomfort, were assessed using the French 18-item questionnaire on discomfort in ICU patients, the “Inconforts des Patients de REAnimation” (IPREA). Patients scored each item from 0 (minimal discomfort) to 10 (maximal discomfort). Associations between patient characteristics at ICU admission, life support therapies and main potentially painful procedures performed during the ICU stay and pain-related discomfort scores assessed at the end of the ICU stay were analyzed.

Results

Patients with complete IPREA questionnaires (n = 2130) were included. The median pain-related discomfort score was 3 (IQR 0–5). From the univariate analysis, pain-related discomfort scores were negatively correlated with age and positively correlated with ICU stay duration; surgical patients reported significant higher pain-related discomfort scores than medical patients; chest drain insertion, chest drain removal, use of bladder catheter, central venous catheter (CVC) insertion, complex dressing change, and intra-hospital transport were associated with pain-related discomfort scores. From the multivariate analyses using generalized estimating equations models, only age, chest drain removal, use of a bladder catheter, CVC insertion, and intra-hospital transport were the main risk factors associated with pain-related discomfort scores.

Conclusion

Patients who underwent chest drain removal, bladder catheter, CVC insertion, and intra-hospital transport during their ICU stay reported higher pain-related discomfort scores (with respect to the whole ICU stay and assessed at the end of their ICU stay) than patients who did not experience these events. This study may pave the way for further targeted studies aiming at investigating a causal link between these common procedures in the ICU and adult critically ill patients’ perceptions of their ICU stay regarding recalled pain.

Trial Registration: Clinicaltrials.gov Identifier NCT02442934, retrospectively registered on May 13, 2015

The 4 U's Rule of Fibromyalgia: A Proposed Model for Fatigue in a Sample of Women with Fibromyalgia: A Qualitative Study

Although fatigue usually goes unnoticed, it is a symptom that poses great challenges to patients with fibromyalgia and is a strong limitation. The aim of this study is to identify and describe the variables involved in fatigue in nine different situations of the Goal Pursuit Questionnaire (GPQ) that may occur in the daily lives of women with fibromyalgia, according to an ABC (Antecedents–Behaviors–Consequences) model. This study followed a qualitative descriptive research method and a deductive–inductive hybrid approach based on a phenomenological paradigm. Twenty-six women with fibromyalgia participated in focus group discussions between February and March of 2018. Thematic content analysis was carried out from transcribed verbatim interviews. We identified nine major themes that emerged from the participants’ conversations: self-imposed duties, muscle fatigue, overwhelming feeling of tiredness, difficulty thinking, difficulty concentrating, negative emotions, lifestyle changes, affected everyday activities, and lack of motivation for daily activities and social interactions. We conclude that the ABC model allowed certain elements to emerge regarding the fatigue experience, highlighting its importance as a symptom in fibromyalgia. This additional analysis of the ABC model showed that fatigue can be described through the 4 U’s Rule, which is integrated by these four adjectives: (1) Unpredictable, (2) Uncontrollable, (3) Unseen, and (4) Unintelligible. Identifying these characteristics can contribute to a better understanding of fibromyalgia in addition to better treatment for these patients.

Sample size calculation in three-level cluster randomized trials using generalized estimating equation models

Three-level cluster randomized trials (CRTs) are increasingly used in implementation science, where 2fold-nested-correlated data arise. For example, interventions are randomly assigned to practices, and providers within the same practice who provide care to participants are trained with the assigned intervention. Teerenstra et al proposed a nested exchangeable correlation structure that accounts for two levels of clustering within the generalized estimating equations (GEE) approach. In this article, we utilize GEE models to test the treatment effect in a two-group comparison for continuous, binary, or count data in three-level CRTs. Given the nested exchangeable correlation structure, we derive the asymptotic variances of the estimator of the treatment effect for different types of outcomes. When the number of clusters is small, researchers have proposed bias-corrected sandwich estimators to improve performance in two-level CRTs. We extend the variances of two bias-corrected sandwich estimators to three-level CRTs. The equal provider and practice sizes were assumed to calculate number of practices for simplicity. However, they are not guaranteed in practice. Relative efficiency (RE) is defined as the ratio of variance of the estimator of the treatment effect for equal to unequal provider and practice sizes. The expressions of REs are obtained from both asymptotic variance estimation and bias-corrected sandwich estimators. Their performances are evaluated for different scenarios of provider and practice size distributions through simulation studies. Finally, a percentage increase in the number of practices is proposed due to efficiency loss from unequal provider and/or practice sizes.

A Pilot Study Evaluating the Effect of Cooler Dialysate Temperature on Hemodynamic Stability During Prolonged Intermittent Renal Replacement Therapy in Acute Kidney Injury

OBJECTIVES

Acute kidney injury requiring renal replacement therapy is associated with high morbidity and mortality. Complications of renal replacement therapy include hemodynamic instability with ensuing shortened treatments, inadequate ultrafiltration, and delay in renal recovery. Studies have shown that lowering dialysate temperature in patients with end-stage renal disease is associated with a decrease in the frequency of intradialytic hypotension. However, data regarding mitigation of hypotension by lowering dialysate temperature in patients with acute kidney injury are scarce. We conducted a prospective, randomized, cross-over pilot study to evaluate the effect of lower dialysate temperature on hemodynamic status of critically ill patients with acute kidney injury during prolonged intermittent renal replacement therapy.

DESIGN

Single-center prospective, randomized, cross-over study.

SETTING

ICUs and a step down unit in a tertiary referral center.

PATIENTS

Acute kidney injury patients undergoing prolonged intermittent renal replacement therapy.

INTERVENTIONS

Participants were randomized to start prolonged intermittent renal replacement therapy with dialysate temperature of 35°C or dialysate temperature of 37°C.

MEASUREMENTS AND MAIN RESULTS

The primary endpoint was the number of hypotensive events, as defined by any of the following: decrease in systolic blood pressure greater than or equal to 20 mm Hg, decrease in mean arterial pressure greater than or equal to 10 mm Hg, decrease in ultrafiltration, or increase in vasopressor requirements. The number of events was analyzed by Poisson regression and other outcomes with repeated-measures analysis of variance. Twenty-one patients underwent a total of 78 prolonged intermittent renal replacement therapy sessions, 39 in each arm. The number of hypotensive events was twice as high during treatments with dialysate temperature of 37°C, compared with treatments with the cooler dialysate (1.49 ± 1.12 vs 0.72 ± 0.69; incidence rate ratio, 2.06; p ≤ 0.0001). Treatment sessions with cooler dialysate were more likely to reach prescribed ultrafiltration targets.

CONCLUSIONS

Patients with acute kidney injury undergoing prolonged intermittent renal replacement therapy with cooler dialysate experienced significantly less hypotension during treatment. Prevention of hemodynamic instability during renal replacement therapy helped to achieve ultrafiltration goals and may help to prevent volume overload in critically ill patients.

Optimal designs in three-level cluster randomized trials with a binary outcome

Cluster randomized trials (CRTs) were originally proposed for use when randomization at the subject level is practically infeasible or may lead to a severe estimation bias of the treatment effect. However, recruiting an additional cluster costs more than enrolling an additional subject in an individually randomized trial. Under budget constraints, researchers have proposed the optimal sample sizes in two-level CRTs. CRTs may have a three-level structure, in which two levels of clustering should be considered. In this paper, we propose optimal designs in three-level CRTs with a binary outcome, assuming a nested exchangeable correlation structure in generalized estimating equation models. We provide the variance of estimators of three commonly used measures: risk difference, risk ratio, and odds ratio. For a given sampling budget, we discuss how many clusters and how many subjects per cluster are necessary to minimize the variance of each measure estimator. For known association parameters, the locally optimal design is proposed. When association parameters are unknown but within predetermined ranges, the MaxiMin design is proposed to maximize the minimum of relative efficiency over the possible ranges, that is, to minimize the risk of the worst scenario.

Assessment of patients' self-perceived intensive care unit discomforts: Validation of the 18-item version of the IPREA

Background and aims

We reported the validation of the 18-item version of the ‘Inconforts des Patients de REAnimation (IPREA)’ questionnaire that includes 2 new items exploring feeling depressed and shortness of breath during an intensive care unit (ICU) stay.

Methods

The validation process was integrated in a multicenter, cluster-randomized, controlled, two-parallel group study built to assess the effectiveness of a tailored multicomponent program for reducing self-perceived discomfort in the ICU. All patients aged 18 years or older who survived an ICU stay of 3 calendar days or more were eligible for inclusion. Data collection included demographics (sex, age), type of admission (medical and surgical), health status scores at admission (Knaus score and McCabe index, Simplified Acute Physiology Score (SAPS) II), specific ICU therapeutics such as mechanical ventilation (MV), noninvasive ventilation (NIV), use of vasopressors, or renal replacement therapy (RRT), and ICU stay duration.

Results

A total of 994 patients were included. The initial structure of IPREA was confirmed using confirmatory factor analysis showing satisfactory fit (RMSEA at 0.042, CFI at 0.912). No multidimensional structure was identified, allowing the calculation of an overall discomfort score. The three highest discomforts were sleep deprivation, thirst, and perfusion lines and other devices, and the 3 lowest discomforts were limited visiting hours, hunger, and isolation. The overall discomfort score of the 18-item version of IPREA did not differ between men and women. Higher age was significantly correlated with a lower overall discomfort score. While MV was not linked to self-reported discomfort, patients treated by NIV reported higher overall discomfort scores than patients not treated by NIV.

Conclusion

The 18-item version of IPREA is easy to use and possesses satisfactory psychometric properties. The availability of a reliable and valid French questionnaire asking about patients’ self-perceived ICU discomforts enables feedback from the health care team to be incorporated in a continuous quality health care improvement strategy.

Trial registration

clinicaltrial.gov NCT02442934 (registration date: May 18, 2015, retrospectively registered).

Tailored multicomponent program for discomfort reduction in critically ill patients may decrease post-traumatic stress disorder in general ICU survivors at 1 year

PURPOSE

Reducing discomfort in the intensive care unit (ICU) should have a positive effect on long-term outcomes. This study assessed whether a tailored multicomponent program for discomfort reduction was effective in reducing post-traumatic stress disorder (PTSD) symptoms at 1 year in general ICU survivors.

METHODS

This study is a prospective observational comparative effectiveness cohort study involving 30 ICUs. It was an extension of the IPREA3 study, a cluster-randomized controlled trial designed to assess the efficacy of a tailored multicomponent program to reduce discomfort in critically ill patients. The program included assessment of ICU-related self-perceived discomforts, immediate and monthly feedback to the healthcare team, and site-specific tailored interventions. The exposure was the implementation of this program. The eligible patients were exposed versus unexposed general adult ICU survivors. The prevalence of substantial PTSD symptoms at 1 year was assessed based on the Impact of Event Scale-Revised (IES-R).

RESULTS

Of the 1537 ICU survivors included in the study, 475 unexposed patients and 344 exposed patients had follow-up data at 1 year: 57 (12.0%) and 21 (6.1%) presented with PTSD at 1 year, respectively (p = 0.004). Considering the clustering and after adjusting for age, gender, McCabe classification, and ICU-related self-perceived overall discomfort score, exposed patients were significantly less likely than unexposed patients to have substantial PTSD symptoms at 1 year (p = 0.015).

CONCLUSIONS

Implementation of a tailored multicomponent program in the ICU that has proved to be effective for reducing self-perceived discomfort in general adult ICU survivors also reduced the prevalence of substantial PTSD symptoms at 1 year.

TRIAL REGISTRATION

ClinicalTrials.gov identifier NCT02762409.

Relative efficiency of unequal versus equal cluster sizes in cluster randomized trials using generalized estimating equation models

There is growing interest in conducting cluster randomized trials (CRTs). For simplicity in sample size calculation, the cluster sizes are assumed to be identical across all clusters. However, equal cluster sizes are not guaranteed in practice. Therefore, the relative efficiency (RE) of unequal versus equal cluster sizes has been investigated when testing the treatment effect. One of the most important approaches to analyze a set of correlated data is the generalized estimating equation (GEE) proposed by Liang and Zeger, in which the "working correlation structure" is introduced and the association pattern depends on a vector of association parameters denoted by ρ. In this paper, we utilize GEE models to test the treatment effect in a two-group comparison for continuous, binary, or count data in CRTs. The variances of the estimator of the treatment effect are derived for the different types of outcome. RE is defined as the ratio of variance of the estimator of the treatment effect for equal to unequal cluster sizes. We discuss a commonly used structure in CRTs-exchangeable, and derive the simpler formula of RE with continuous, binary, and count outcomes. Finally, REs are investigated for several scenarios of cluster size distributions through simulation studies. We propose an adjusted sample size due to efficiency loss. Additionally, we also propose an optimal sample size estimation based on the GEE models under a fixed budget for known and unknown association parameter (ρ) in the working correlation structure within the cluster.

A tailored multicomponent program to reduce discomfort in critically ill patients: a cluster-randomized controlled trial

PURPOSE

Critically ill patients are exposed to stressful conditions and experience several discomforts. The primary objective was to assess whether a tailored multicomponent program is effective for reducing self-perceived discomfort.

METHODS

In a cluster-randomized two-arm parallel trial, 34 French adult intensive care units (ICUs) without planned interventions to reduce discomfort were randomized, 17 to the arm including a 6-month period of program implementation followed by a 6-month period without the program (experimental group), and 17 to the arm with an inversed sequence (control group). The tailored multicomponent program consisted of assessment of ICU-related self-perceived discomforts, immediate and monthly feedback to healthcare teams, and site-specific tailored interventions. The primary outcome was the overall discomfort score derived from the 16-item IPREA questionnaire (0, minimal, 100, maximal overall discomfort) and the secondary outcomes were the discomfort scores of each IPREA item. IPREA was administered on the day of ICU discharge with a considered timeframe from the ICU admission until ICU discharge.

RESULTS

During a 1-month assessment period, 398 and 360 patients were included in the experimental group and the control group, respectively. The difference (experimental minus control) of the overall discomfort score between groups was - 7.00 (95% CI - 9.89 to - 4.11, p < 0.001). After adjustment (age, gender, ICU duration, mechanical ventilation duration, and type of admission), the program effect was still positive for the overall discomfort score (difference - 6.35, SE 1.23, p < 0.001) and for 12 out of 16 items.

CONCLUSIONS

This tailored multicomponent program decreased self-perceived discomfort in adult critically ill patients.

TRIAL REGISTRATION

Clinicaltrials.gov Identifier NCT02442934.

Estimation of critically ill patients' complaints by the nurse, the physician and the patient's family: A prospective comparative study

OBJECTIVES

To evaluate the ability of the families of critically ill patients and of the intensive care team caring for the patient to communicate and accurately identify patients' complaints.

DESIGN

The complaints of critically ill patients were evaluated by a psychologist using a list of 12 items. The same day as the patient interview, the psychologist collected an estimation of the patient's complaints from the family, the nurse and the physician.

SETTING

20-bed Intensive Care Unit in a large University Hospital.

MAIN OUTCOME MEASURES

Patients' complaints.

RESULTS

51 patients were included. The most frequently reported complaints were insomnia, the inability to talk and presence of a tracheal tube. Patients reported a significantly higher prevalence of "misunderstanding" than that estimated by the nurses (55% vs 33%, p=0.045). The reported prevalence of "inability to talk" as the main complaint was significantly higher among patients than estimated by nurses and physicians (16% vs 2%, p=0.03 and 16% vs 2%, p=0.03 respectively). For the analysis of the individual complaints, there was a poor agreement between the patients and the other respondents.

CONCLUSION

This study found that the estimation of critically ill patients' complaints by their families, nurses and physicians was largely suboptimal.

Optimal design of longitudinal data analysis using generalized estimating equation models

Longitudinal studies are often applied in biomedical research and clinical trials to evaluate the treatment effect. The association pattern within the subject must be considered in both sample size calculation and the analysis. One of the most important approaches to analyze such a study is the generalized estimating equation (GEE) proposed by Liang and Zeger, in which "working correlation structure" is introduced and the association pattern within the subject depends on a vector of association parameters denoted by ρ. The explicit sample size formulas for two-group comparison in linear and logistic regression models are obtained based on the GEE method by Liu and Liang. For cluster randomized trials (CRTs), researchers proposed the optimal sample sizes at both the cluster and individual level as a function of sampling costs and the intracluster correlation coefficient (ICC). In these approaches, the optimal sample sizes depend strongly on the ICC. However, the ICC is usually unknown for CRTs and multicenter trials. To overcome this shortcoming, Van Breukelen et al. consider a range of possible ICC values identified from literature reviews and present Maximin designs (MMDs) based on relative efficiency (RE) and efficiency under budget and cost constraints. In this paper, the optimal sample size and number of repeated measurements using GEE models with an exchangeable working correlation matrix is proposed under the considerations of fixed budget, where "optimal" refers to maximum power for a given sampling budget. The equations of sample size and number of repeated measurements for a known parameter value ρ are derived and a straightforward algorithm for unknown ρ is developed. Applications in practice are discussed. We also discuss the existence of the optimal design when an AR(1) working correlation matrix is assumed. Our proposed method can be extended under the scenarios when the true and working correlation matrix are different.