Skip to main content
Download PDF

Statistical analysis plan for a cluster-randomized crossover trial comparing the effectiveness and safety of a flexible family visitation model for delirium prevention in adult intensive care units (the ICU Visits Study)

Trials volume 19, Article number: 636 (2018) Cite this article

Abstract

Background

Most adult intensive care units (ICUs) worldwide adopt restrictive family visitation models (RFVMs). However, evidence, mostly from non-randomized studies, suggests that flexible adult ICU visiting hours are safe policies that can result in benefits such as prevention of delirium and increase in satisfaction with care. Accordingly, the ICU Visits Study was designed to compare the effectiveness and safety of a flexible family visitation model (FFVM) vs. an RFVM on delirium prevention among ICU patients, and also to analyze its potential effects on family members and ICU professionals.

Methods/design

The ICU Visits Study is a cluster-randomized crossover trial which compares an FFVM (12 consecutive ICU visiting hours per day) with an RFVM (< 4.5 ICU visiting hours per day) in 40 Brazilian adult ICUs. Participant ICUs are randomly assigned to either an FFVM or RFVM in a 1:1 ratio. After enrollment and follow-up of 25 patients, each ICU is crossed over to the other visitation model, until 25 more patients per site are enrolled and followed. The primary outcome is the cumulative incidence of delirium measured by the Confusion Assessment Method for the ICU. Secondary and tertiary outcomes include relevant measures of effectiveness and safety of ICU visiting policies among patients, family members, and ICU professionals. Herein, we describe all primary statistical procedures that will be used to evaluate the results and perform exploratory and sensitivity analyses of this study. This pre-specified statistical analysis plan was written and submitted without knowledge of the study data.

Discussion

This a priori statistical analysis plan aims to enhance the transparency of our study, facilitating unbiased analyses of ICU visit study data, and provide guidance for statistical analysis for groups conducting studies in the same field.

Trial registration

ClinicalTrials.gov, NCT02932358. Registered on 11 October 2016.

Peer Review reports

Background

The recognition of the important role of family members in the intensive care unit (ICU) is the cornerstone of patient-centered care [1]. Beyond the justification of humanization, this strategy is proposed as a means to improve patient and family member outcomes [2]. In this context, small studies with a predominant before and after design have shown an association between flexible ICU visiting hours and reduced incidence of delirium in patients and increased satisfaction with care in family members [3,4,5]. Despite the growing recognition of the importance of family presence in the ICU, most ICUs around the world still adopt restricted visiting hours motivated by the ICU professionals’ perceptions of increased risk of disorganization of care, infection transmission, and burnout [6,7,8,9]. Unfortunately, few randomized studies have focused on the evaluation of the impact of different ICU visiting models on patients, family members, and ICU professionals, and this lack of evidence may constitute a barrier to the implementation of patient-centered care interventions in the ICU. The ICU Visits Study aims to investigate the effectiveness and safety of a flexible family visitation model (FFVM) vs. a restrictive family visitation model (RFVM) on delirium prevention among ICU patients, and to evaluate its potential benefits and hazards for family members and ICU professionals.

The present statistical analysis plan (SAP) aims to describe the trial’s analytical objectives and procedures before the end of the study recruitment and locking of the trial database to start analyses to comply with good clinical practice and avoid outcome reporting bias. This SAP was drafted without knowledge of any of the results of the investigators.

Trial overview

The ICU Visits Study is a cluster-randomized, crossover trial comparing an FFVM (12 consecutive ICU visiting hours per day) with an RFVM (< 4.5 ICU visiting hours per day) in Brazilian adult ICUs. The study background, design, rationale, eligibility criteria, and sample size have been previously published [10]. In brief, mixed adult ICUs of public and private philanthropic hospitals with a restrictive policy of ICU visiting hours (< 4.5 h/day) are randomly assigned to either an FFVM or RFVM in a 1:1 ratio until the recruitment and follow-up of 25 patients (phase 1) (Fig. 1). After a 30-day washout period without subject recruitment, each ICU is switched over to the other visitation model (phase 2) until 25 more patients per site are enrolled and followed. The randomization is stratified by the number of ICU beds (1 to 10 or > 10 ICU beds) and performed using random block sizes of 2, 4, and 6. Consecutive patients aged ≥18 years admitted to the ICU, their closest family members, and bedside ICU professionals in each cluster are considered eligible for the present trial. A complete description of the inclusion and exclusion criteria is shown in Table 1.

Fig. 1
figure 1

Study design. FFVM flexible family visitation model, ICU intensive care unit, RFVM restrictive family visitation model. All ICUs will have a learning period within the first 15 days of phases 1 and 2. During this period, ICUs will receive the intervention (FFVM or RFVM) but will not recruit subjects. Local investigators will use this period to adapt the ICU staff to the organizational aspects of study interventions

Full size image
Table 1 Eligibility criteria
Full size table

In the FFVM, patients are allowed to receive visits from one or two close family members for up to 12 consecutive hours each day. Family members who agree to join the family visits have to attend a structured meeting at the ICU in which they receive guidance about the ICU environment, common ICU treatments, rehabilitation and basic infection control practices, multidisciplinary work at the ICU, and information on palliative care and delirium prevention. Additionally, family members receive an information brochure and are encouraged to access a website, both of which are designed to explain what happens during and after an ICU stay to legitimize emotions and improve cooperation with relatives without increasing the ICU staff workload. In addition to family visitation, patients in the FFVM are allowed to receive social visits at specific time intervals (according to the local ICU policies) from friends or other family members who did not qualify for flexible family visitation. In the RFVM, patients are allowed visitors according to routine ICU practices, but limited to the maximum of 4.5 h of visitation per day. Visitors are not required to attend the structured meeting in the RFVM.

The study primary outcome is the cumulative incidence of delirium during the ICU stay measured by trained researchers using the Confusion Assessment Method for the ICU (CAM-ICU) [11]. Secondary outcomes include daily hazard of delirium, ventilator-free days at day 7, any ICU-acquired infections according to the Centers for Disease Control and Prevention (CDC) criteria [12,13,14], ICU length of stay, and hospital mortality among the patients; symptoms of anxiety and depression measured by the Hospital Anxiety and Depression Scale (HADS) [15] and satisfaction measured by the Critical Care Family Needs Inventory (CCFNI) [16] among the family members; and prevalence of burnout syndrome evaluated by the Maslach Burnout Inventory (MBI) [17] among the ICU professionals. Tertiary outcomes include need for antipsychotic agents, need for mechanical restraints, unplanned loss of invasive devices (venous catheter, enteral tube, or urinary catheter), coma-free days at day 7, and ICU-acquired pneumonia, urinary tract infection, or bloodstream infection among the patients; self-perception of involvement in patient care among the family members; and satisfaction among the ICU professionals.

Statistical analysis plan

Overall principles

The main analysis for each outcome will be performed at the subject level using the intention-to-treat (ITT) principle, meaning that all participants with a recorded outcome will be included in the analysis and will be analyzed according to the treatment group to which they were randomized (FFVM or RFVM), independent of actual visiting hours. Moreover, all analyses will account for the cluster-randomized crossover design to ensure correct type I error rates and confidence intervals (CIs). A significance level of 0.05, adjusted for multiple comparisons when appropriate, will be used for all statistical comparisons. Analysis will start once all data to discharge for the last included patient have been obtained, the database has been cleaned and locked, and the SAP has been submitted for publication. The R Development Core Team software will be used for analysis [18].

Handling of missing data

We anticipate minimal missing values, given that the study variables will be determined during hospital stay by trained researchers. Nevertheless, the coordinator center will contact site investigators to retrieve any missing data values.

ICUs with incomplete subject recruitment (e.g., less than 50 patients) will be included in the primary analysis for the study outcomes considering all subjects available in the cluster. To assess the risk of bias related to ICUs that did not achieve the patient recruitment goal, multiple imputation techniques will be performed for the primary outcome and presented as sensitivity analyses. The missing values for the variables that compose the HADS, CCFNI, and MBI will be imputed, replacing the missing items with the mean of the answered items in the same subscale, if at least half of that subscale has been answered. The missing values for the variables that compose the PREdiction of DELIRium in ICU patients (PRE-DELIRIC) score will be imputed in a similar way as in the original study [19]. We will assume that if a variable is not determined, most likely the missing variable has a normal or negative value (e.g., no infection, no metabolic acidosis) or a mean value (e.g., Acute Physiology and Chronic Health Evaluation-II [APACHE-II] score) of the study population. We will not perform any imputation for the length of ICU visits per patient per day.

Definition of analysis sets

At the cluster level, the ITT population includes all randomized ICUs that recruited subjects, regardless of the degree of adherence to the study interventions or achievement of the patient recruitment goal (50 patients per ICU). At the subject level, the ITT population includes all participants, regardless of protocol deviations. This includes patients who did not receive any ICU visits and family members who did not visit patients during the ICU stay, as well as patients who received visits longer than the maximum limit of visiting hours in both study periods (FFVM and RFVM).

Statistical analyses

Patient flow

The flow of participants will be displayed in accordance with the Consolidated Standards of Reporting Trials (CONSORT) flow diagram (CONSORT 2010 statement: extension to cluster-randomized trials) [20]. This description will include information about eligibility criteria and follow-up losses at both cluster and subject levels.

Adherence to study interventions

The total length of visits per patient per day will be evaluated as the primary adherence variable to the study interventions. We will consider the length of stay of all visitors (whether family members or not) at the bedside of an enrolled patient during the ICU stay. For analytical purposes, we will not consider overlapping visiting hours in the total length of visits per patient per day (i.e., only the period in which there was at least one visitor [apart from the number of visitors] will be considered). The differences in the means of length of visits per patient per day between the FFVM and the RFVM groups will be compared using generalized estimating equations with adjustment for the cluster effect, period effect, and interaction between cluster and period. Additionally, we will describe the FFVM family members’ adherence to the website, and the healthcare professionals’ perception of effectiveness and barriers for implementation of the FFVM.

To assess the fidelity of FFVM implementation, we will perform on-site monitoring visits in all participant ICUs. During these visits, the ICU staff perception about adherence to proposed FFVM processes will be assessed by the coordinating center researchers using semi-structured interviews. During the FFVM period, each ICU will be rated from 0 to 100% (with higher percentages indicating higher levels of adherence to intended FFVM processes) in the following domains: (1) Visiting hours - adherence to visiting hours according to the study protocol; (2) Dissemination - dissemination of the FFVM to family members of ICU patients; (3) Structured meetings - frequency and quality of structured meetings; (4) Staff training - education of the ICU staff about the FFVM procedures. The total fidelity of the FFVM implementation score represents the mean of the four evaluated domains. We plan to perform sensitivity and exploratory analyses considering the effects of FFVM implementation on outcomes.

Baseline characteristics

The baseline characteristics of all participants will be presented by study arm in a table (see Table 2), but no formal statistical hypothesis testing will be performed to avoid unnecessary testing.

Table 2 Baseline characteristics of study participants
Full size table

Inter-rater reliability of the CAM-ICU

Inter-rater reliability measures of the CAM-ICU will be performed before study initiation to evaluate the quality of assessment. In each participant ICU, we will assess the concordance and agreement in diagnostic classification between trained intensive care physicians (reference diagnosis) and local outcome evaluators.

Primary outcome

All the pre-specified study outcomes are described in Table 3. The primary outcome is the cumulative incidence of delirium during ICU stay determined by the CAM-ICU, which was validated for the Brazilian population of critical care patients [21]. The cumulative incidence of delirium is defined as the presence of delirium (at least one positive CAM-ICU score) during the ICU stay. The differences in the incidences of delirium between the FFVM and the RFVM groups will be compared using generalized estimating equations with adjustment for the cluster effect, period effect, and interaction between intervention and period and presented as risk ratio (RR) and 95% CI.

Table 3 Study outcomes
Full size table

Sensitivity analyses for the primary outcome

We aim to conduct the following sensitivity analyses for the primary outcome to assess consistency and the risk of bias:

  1. 1.

    Evaluation of study intervention effects adjusted by the baseline risk of delirium according to the PRE-DELIRIC score.

  2. 2.

    Evaluation of study intervention effects considering the cluster adherence to FFVM implementation. In this analysis, the study intervention effects will be adjusted by the total fidelity of FFVM implementation score.

  3. 3.

    Evaluation of study intervention effects considering the potential confounding effect of sedation on the delirium diagnosis. In this analysis, the study intervention effects will be evaluated, considering patients with positive CAM-ICU during ICU stay in the context of Richmond Agitation-Sedation Scale (RASS) score – 2 or − 3 as non-delirious subjects.

  4. 4.

    Evaluation of effects of study interventions, considering a best plausible scenario/worst plausible scenario imputation of outcomes among ICUs that did not complete the patient recruitment as follows:

    1. a)

      Best plausible scenario: Imputation of the lowest study cluster/phase incidence of delirium for missing FFVM patients and the highest study cluster/phase incidence of delirium for missing RFVM patients

    2. b)

      Worst plausible scenario: Imputation of the lowest study cluster/phase incidence of delirium for missing RFVM patients and the highest study cluster/phase incidence of delirium for missing FFVM patients.

Subgroup analyses for the primary outcome

There will be three a priori defined subgroup analyses for the primary endpoint: (1) effectiveness of the FFVM vs. RFVM in ICUs according to the PRE-DELIRIC score (patients with a predicted risk < 25%, 25–50%, 50–75%, and > 75%); (2) effectiveness of the FFVM vs. RFVM according to patient group (medical vs. surgical and neurocritical vs. non-neurocritical); and (3) effectiveness of the FFVM vs. RFVM in ICUs according to APACHE-II scores (≤15 vs. > 15 points). The consistency of intervention effects across the above-mentioned subgroups will be assessed by means of tests for interaction. The Bonferroni correction will be applied to adjust the subgroup analyses for multiple comparisons.

Secondary outcomes

We will use the following statistical procedures to evaluate the study secondary outcomes:

  1. 1.

    Daily hazard of delirium: The daily hazard of delirium will be analyzed using a joint survival model that accounts for the treatment effect (with adjustment for the cluster effect, period effect, and interaction between cluster and period) on repeated daily indicator of delirium within each patient and terminating event (death or discharge from the ICU) [22]. This outcome will be presented as hazard ratio (HR) and 95% CI.

  2. 2.

    Any ICU-acquired infections: The differences in the cumulative incidence of any ICU-acquired infection between the FFVM and RFVM groups will be compared using generalized estimating equations with adjustment for the cluster effect, period effect, and interaction between intervention and period. This outcome will be presented as RR and 95% CI.

  3. 3.

    Ventilator-free days: The number of ventilator-free days will be evaluated using the 7-day time horizon. Ventilator-free days will be set to 0 for patients who have died. The differences in the mean ventilator-free days between the FFVM and RFVM groups will be compared using generalized estimating equations with adjustment for the cluster effect, period effect, and interaction between intervention and period. This outcome will be presented as mean difference (MD) and 95% CI.

  4. 4.

    ICU length of stay: The differences in the mean ICU length of stay between the FFVM and RFVM groups will be compared using generalized estimating equations with adjustment for the cluster effect, period effect, and interaction between intervention and period. This outcome will be presented as MD and 95% CI.

  5. 5.

    Hospital mortality: The differences in hospital mortality rates between the FFVM and RFVM groups will be compared using generalized estimating equations with adjustment for the cluster effect, period effect, and interaction between intervention and period. This outcome will be presented as RR and 95% CI.

  6. 6.

    Symptoms of anxiety and depression among family members: The differences in the mean HADS scores for anxiety and depression between the FFVM and RFVM groups will be compared using generalized estimating equations with adjustment for the cluster effect, period effect, and interaction between intervention and period. These outcomes will be presented as MD and 95% CI.

  7. 7.

    Satisfaction among family members: The differences in the mean CCFNI satisfaction scores between the FFVM and RFVM groups will be compared using generalized estimating equations with adjustment for the cluster effect, period effect, and interaction between intervention and period. This outcome will be presented as MD and 95% CI.

  8. 8.

    Prevalence of burnout among the ICU professionals: Burnout is defined as a cumulative MBI score greater than − 9. The differences in the prevalence of burnout between the FFVM and RFVM groups will be compared using generalized estimating equations with adjustment for the cluster effect and baseline MBI total scores. This outcome will be presented as prevalence ratio and 95% CI.

Given that the main safety outcomes of the present study are any ICU-acquired infections and burnout, we will not adjust these outcomes for multiple comparisons. For other secondary outcomes, the Bonferroni correction will be applied to adjust the analyses for multiple comparisons when appropriate, taking into consideration the number of comparisons within each population of interest (patients, family members, and ICU professionals).

We will perform sensitivity analyses to check the consistency of intervention effects on family members and ICU professionals as follows:

  1. 1.

    Evaluation of study intervention effects on HADS subscale scores among family members, considering HADS scores as categorical variables with established cutoff points (> 10 points for both anxiety and depression subscales)

  2. 2.

    Evaluation of study intervention effects on HADS subscale scores among family members adjusted by previous history of anxiety (for anxiety HADS subscale) and depression (for depression HADS subscale)

  3. 3.

    Evaluation of study intervention effects on CCFNI subscale domains (information, proximity, reassurance, support, and comfort)

  4. 4.

    Evaluation of study intervention effects on incidence (excluding individuals with burnout at baseline) and prevalence of burnout among ICU professionals, considering the following alternative MBI criteria:

    1. a)

      Alternative criteria 1: An emotional exhaustion subscale score ≥ 27, OR a depersonalization subscale score ≥ 10, OR a personal accomplishment subscale score ≤ 33

    2. b)

      Alternative criteria 2: An emotional exhaustion subscale score ≥ 27, OR a depersonalization subscale score ≥ 10

    3. c)

      Alternative criteria 3: An emotional exhaustion subscale score ≥ 27, AND a depersonalization subscale score ≥ 10, AND a personal accomplishment subscale score ≤ 33.

Tertiary outcomes

We will use the following statistical procedures to evaluate the study tertiary outcomes:

  1. 1.

    Need for antipsychotic agents: The differences in the cumulative incidence of need for antipsychotic agents during ICU stay between the FFVM and RFVM groups will be compared using generalized estimating equations with adjustment for the cluster effect, period effect, and interaction between intervention and period. This outcome will be presented as RR and 95% CI.

  2. 2.

    Need for mechanical restraints: The differences in the cumulative incidence of need for mechanical restraints during ICU stay between the FFVM and RFVM groups will be compared using generalized estimating equations with adjustment for the cluster effect, period effect, and interaction between intervention and period. This outcome will be presented as RR and 95% CI.

  3. 3.

    Unplanned loss of invasive devices: The differences in the cumulative incidence of any unplanned loss of invasive devices (venous catheter, enteral tube, or urinary catheter) during the ICU stay between the FFVM and RFVM groups will be compared using generalized estimating equations with adjustment for the cluster effect, period effect, and interaction between intervention and period. This outcome will be presented as RR and 95% CI.

  4. 4.

    Coma-free days at day 7: The number of coma-free days will be evaluated using the 7-day time horizon. Coma-free days will be set to 0 for patients who have died. The differences in the mean proportion of days free of coma (RASS [23] – 4 or − 5) between the FFVM and RFVM groups will be compared using generalized estimating equations with adjustment for the cluster effect, period effect, and interaction between intervention and period. This outcome will be presented as MD and 95% CI.

  5. 5.

    ICU-acquired pneumonia: The differences in the cumulative incidence of ICU-acquired pneumonia between the FFVM and RFVM groups will be compared using generalized estimating equations with adjustment for the cluster effect, period effect, and interaction between intervention and period. This outcome will be presented as RR and 95% CI.

  6. 6.

    ICU-acquired urinary tract infection: The differences in the cumulative incidence of urinary tract infection between the FFVM and RFVM groups will be compared using generalized estimating equations with adjustment for the cluster effect, period effect, and interaction between intervention and period. This outcome will be presented as RR and 95% CI.

  7. 7.

    ICU-acquired bloodstream infection: The differences in the cumulative incidence of bloodstream infection between the FFVM and RFVM groups will be compared using generalized estimating equations with adjustment for the cluster effect, period effect, and interaction between intervention and period. This outcome will be presented as RR and 95% CI.

  8. 8.

    Self-perception of involvement in patient care among the family members: The self-perception of involvement in patient care will be evaluated using a score developed for the present study, which is composed of 9 questions related to the self-perception of involvement of the family member in the care of the patient in the following domains: (1) re-orientation activities, (2) pain control, (3) mobilization, (4) feeding, (5) hygiene, (6) emotional support, (7) helping patients to interpret ICU staff orientations, (8) helping the ICU professionals to understand patient needs, and (9) helping to create a patient-friendly environment in the patient room. The score of each question ranges from 0 (never involved in the activity) to 3 (very frequently involved in the activity). The total score, which varies from 0 (no involvement) to 27 (maximum degree of involvement), will be obtained by the sum of the scores of each question. The differences in the mean total scores of self-perception of involvement in patient care between the FFVM and RFVM groups will be compared using generalized estimating equations with adjustment for the cluster effect, period effect, and interaction between intervention and period. This outcome will be presented as MD and 95% CI.

  9. 9.

    Satisfaction among ICU professionals: The satisfaction of ICU professionals will be evaluated through the score of the following question: Are you satisfied with the current visiting policy of your ICU? The responses to this 5-option Likert scale question may be one of the following: 0 - unsatisfied, 1- somewhat dissatisfied, 2 - indifferent, 3 - somewhat satisfied, 4 - very satisfied. The differences in the mean score values between the FFVM and RFVM groups will be compared using generalized estimating equations with adjustment for the cluster effect. This outcome will be presented as MD and 95% CI.

No adjustment for multiple comparisons will be made for tertiary outcomes. Therefore, the results of tertiary outcomes should be considered exploratory.

Differences between the study protocol and statistical analysis plan

There are no differences between the study protocol and this SAP in relation to the proposed outcomes or statistical procedures.

Discussion and trial status

In this SAP, we present the statistical procedures that will allow the comparison of effectiveness and safety outcomes between the FFVM and RFVM in the ICU Visits Study. The present publication aims to avoid risks of outcome reporting bias and data-driven results and provide guidance for statistical analysis for future studies in this field. As of March 2018, 40 ICUs were randomized. Currently, 1591 patients, 1192 family members, and 829 ICU professionals were included in the study. We expect that the recruitment of subjects will be completed in June 2018.

Abbreviations

APACHE-II:

Acute Physiology and Chronic Health Evaluation-II

BSI:

Bloodstream infection

CAM-ICU:

Confusion Assessment Method for the Intensive Care Unit

CCFNI:

Critical Care Family Needs Inventory

CDC:

Centers for Disease Control and Prevention

CI:

Confidence interval

CONSORT:

Consolidated Standards of Reporting Trials

FFVM:

Flexible family visitation model

HADS:

Hospital Anxiety and Depression Scale

HR:

Hazard ratio

ICU:

Intensive care unit

IQR:

Interquartile range

MBI:

Maslach Burnout Inventory

MD:

Mean difference

PR:

Prevalence ratio

PRE-DELIRIC:

PREdiction of DELIRium in ICU patients

RASS:

Richmond Agitation-Sedation Scale

RFVM:

Restrictive family visitation model

RR:

Risk ratio

SAP:

Statistical analysis plan

SD:

Standard deviation

SOFA:

Sequential Organ Failure Assessment

USD:

United States dollars

UTI:

Urinary tract infection

References

  1. Gerritsen RT, Hartog CS, Curtis JR. New developments in the provision of family-centered care in the intensive care unit. Intensive Care Med. 2017;43:550–3.

    Article  Google Scholar 

  2. Goldfarb MJ, Bibas L, Bartlett V, Jones H, Khan N. Outcomes of patient- and family-centered care interventions in the ICU: a systematic review and meta-analysis. Crit Care Med. 2017;45:1751–61.

    Article  Google Scholar 

  3. Rosa RG, Tonietto TF, da Silva DB, Gutierres FA, Ascoli AM, Madeira LC, et al. Effectiveness and safety of an extended ICU visitation model for delirium prevention: a before and after study. Crit Care Med. 2017;45:1660–7.

    Article  Google Scholar 

  4. Westphal GA, Moerschberger MS, Vollmann DD, Inácio AC, Machado MC, Sperotto G, et al. Effect of a 24-h extended visiting policy on delirium in critically ill patients. Intensive Care Med. 2018;44:968–70. https://doi.org/10.1007/s00134-018-5153-5 [Epub ahead of print].

    Article  PubMed  Google Scholar 

  5. Nassar Junior AP, Besen BAMP, Robinson CC, Falavigna M, Teixeira C, Rosa RG. Flexible versus restrictive visiting policies in ICUs: a systematic review and meta-analysis. Crit Care Med. 2018;46:1175–80. https://doi.org/10.1097/CCM.0000000000003155 [Epub ahead of print].

    Article  PubMed  Google Scholar 

  6. Ramos FJ, Fumis RR, de Azevedo LC, Schettino G. Intensive care unit visitation policies in Brazil: a multicenter survey. Rev Bras Ter Intensiva. 2014;26:339–46.

    Article  Google Scholar 

  7. Garrouste-Orgeas M, Vinatier I, Tabah A, Misset B, Timsit JF. Reappraisal of visiting policies and procedures of patient’s family information in 188 French ICUs: a report of the Outcomerea Research Group. Ann Intensive Care. 2016;6:82.

    Article  Google Scholar 

  8. Noordermeer K, Rijpstra TA, Newhall D, Pelle AJ, van der Meer NJ. Visiting policies in the adult intensive care units in the Netherlands: survey among ICU directors. ISRN Critical Care. 2013;137045:1–6.

    Article  Google Scholar 

  9. Lee MD, Friedenberg AS, Mukpo DH, Conray K, Palmisciano A, Levy MM. Visiting hours policies in New England intensive care units: strategies for improvement. Crit Care Med. 2007;35:497–501.

    Article  CAS  Google Scholar 

  10. Rosa RG, Falavigna M, Robinson CC, da Silva DB, Kochhann R, de Moura RM, et al. Study protocol to assess the effectiveness and safety of a flexible family visitation model for delirium prevention in adult intensive care units: a cluster-randomised, crossover trial (The ICU Visits Study). BMJ Open. 2018;8:e021193.

    Article  Google Scholar 

  11. Ely EW, Margolin R, Francis J, May L, Truman B, Dittus R, et al. Evaluation of delirium in critically ill patients: validation of the Confusion Assessment Method for the Intensive Care Unit (CAM-ICU). Crit Care Med. 2001;29:1370–9.

    Article  CAS  Google Scholar 

  12. Centers for Disease Control and Prevention. Bloodstream Infection Event (Central Line-Associated Bloodstream Infection and non-central line-associated Bloodstream Infection). 2017. https://www.cdc.gov/nhsn/pdfs/pscmanual/4psc_clabscurrent.pdf. Accessed 14 Jun 2017.

  13. Centers for Disease Control and Prevention. Pneumonia (Ventilator-associated [VAP] and non-ventilator-associated Pneumonia [PNEU]) Event. 2017. https://www.cdc.gov/nhsn/pdfs/pscmanual/6pscvapcurrent.pdf. Accessed 14 Jun 2017.

  14. Centers for Disease Control and Prevention. Urinary Tract Infection (Catheter-Associated Urinary Tract Infection [CAUTI] and Non-Catheter-Associated Urinary Tract Infection [UTI]) and Other Urinary System Infection [USI] Events. 2017. https://www.cdc.gov/nhsn/pdfs/pscManual/7pscCauticurrent.pdf. Accessed 14 Jun 2017.

    Google Scholar 

  15. Botega NJ, Bio MR, Zomignani MA, Garcia C Jr, Pereira WA. Mood disorders among medical in-patients: a validation study of the hospital anxiety and depression scale (HAD). Rev Saude Publica. 1995;29:355–63.

    Article  CAS  Google Scholar 

  16. Fumis RR, Nishimoto IN, Deheinzelin D. Measuring satisfaction in family members of critically ill cancer patients in Brazil. Intensive Care Med. 2006;32:124–8.

    Article  Google Scholar 

  17. Campos JA, Maroco J. Maslach Burnout Inventory - Student Survey: Portugal-Brazil cross-cultural adaptation. Rev Saude Publica. 2012;46:816–24.

    Article  Google Scholar 

  18. R Development Core Team. R: A language and environment for statistical computing. Vienna: R Foundation for Statistical Computing; 2016. https://www.R-project.org/. Accessed 14 Jun 2017

    Google Scholar 

  19. van den Boogaard M, Pickkers P, Slooter AJ, Kuiper MA, Spronk PE, van der Voort PH, et al. Development and validation of PRE-DELIRIC (PREdiction of DELIRium in ICu patients) delirium prediction model for intensive care patients: observational multicentre study. BMJ. 2012;344:e420.

    Article  CAS  Google Scholar 

  20. Campbell MK, Piaggio G, Elbourne DR, Altman DG. Consort 2010 statement: extension to cluster randomised trials. BMJ. 2012;345:e5661.

    Article  Google Scholar 

  21. Gusmao-Flores D, Salluh JI, dal-Pizzol F, Santana LR, Lins RM, Lemos PP, et al. Validity and reliability of the Brazilian-Portuguese version of three tools to diagnose delirium: CAM-ICU, CAM-ICU Flowsheet and ICDSC. Crit Care. 2011;15(Suppl 2):P50.

    Article  Google Scholar 

  22. Colantuoni E, Dinglas VD, Ely EW, Hopkins RO, Needham DM. Statistical methods for evaluating delirium in the ICU. Lancet Repir Med. 2016;4(7):534–6.

    Article  Google Scholar 

  23. Sessler CN, Gosnell MS, Grap MJ, Brophy GM, O'Neal PV, Keane KA, et al. The Richmond Agitation-Sedation Scale: validity and reliability in adult intensive care unit patients. Am J Respir Crit Care Med. 2002;166:1338–44.

    Article  Google Scholar 

Download references

Acknowledgements

The authors thank Hospital Moinhos de Vento and the Brazilian Ministry of Health for their support in conducting the study.

Funding

The present study was funded by the Brazilian Ministry of Health through the Program of Institutional Development of the Brazilian Unified Health System (PROADI-SUS). The trial is coordinated by Hospital Moinhos de Vento and endorsed by the Brazilian Research in Intensive Care Network (BRICNet).

Availability of data and materials

The datasets used and/or analyzed during the current study will be available from the corresponding author on reasonable request.

Author information

Authors and Affiliations

  1. Research Projects Office, Hospital Moinhos de Vento (HMV), Rua Ramiro Barcelos, 910, Moinhos de Vento, Porto Alegre, RS, 90035-001, Brazil

    Daniel Sganzerla, Caroline Cabral Robinson, Renata Kochhann, Mariana Martins Siqueira Santos, Rafaela Moraes de Moura, Mirceli Goulart Barbosa, Daniel Schneider, Débora Mariani & Rodrigo Wiltgen Jeffman

  2. Intensive Care Unit, HMV. Rua Ramiro Barcelos, 910, Moinhos de Vento, Porto Alegre, RS, 90035-001, Brazil

    Cassiano Teixeira, Daiana Barbosa da Silva, Tarissa Ribeiro, Cláudia Eugênio & Regis Goulart Rosa

  3. Department of Critical Care, Instituto D’Or de Pesquisa e Ensino (IDOR), Rua Diniz Cordeiro, 30, Botafogo, Rio de Janeiro, RJ, 22281-100, Brazil

    Fernando Bozza

  4. HCor Research Institute, Rua Abílio Soares, 250, Paraíso, São Paulo, SP, 04005-909, Brazil

    Alexandre Biasi Cavalcanti, Jorge Ibrain Salluh & Lucas Petri Damiani

  5. Intensive Care Unit, Hospital Sírio-Libanês, Rua Dona Adma Jafet, 91, Bela Vista, São Paulo, SP, 01308-050, Brazil

    Luciano Cesar Pontes Azevedo

  6. Department of Anesthesiology, Pain and Intensive Care, Universidade Federal de São Paulo (UNIFESP), Rua Napoleão de Barros 737, Vila Clementino, São Paulo, SP, 04024-900, Brazil

    Flávia Ribeiro Machado

  7. Intensive Care Unit, Hospital de Clínicas de Porto Alegre (HCPA), Rua Ramiro Barcelos, 2350, Santa Cecília, Porto Alegre, RS, 90035-903, Brazil

    José Augusto Santos Pellegrini

  8. Department of Internal Medicine, School of Medicine, Universidade Federal de Ciências da Saúde de Porto Alegre (UFCSPA), Rua Sarmento Leite, 245, Centro Histórico, Porto Alegre, RS, 90050-170, Brazil

    Nilton Brandão da Silva

  9. Institute for Education and Research, HMV, Rua Ramiro Barcelos, 910, Moinhos de Vento, Porto Alegre, RS, 90035-001, Brazil

    Rafael Barberena Moraes & Maicon Falavigna

Authors
  1. Daniel Sganzerla
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Cassiano Teixeira
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Caroline Cabral Robinson
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Renata Kochhann
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Mariana Martins Siqueira Santos
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Rafaela Moraes de Moura
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Mirceli Goulart Barbosa
    View author publications

    You can also search for this author in PubMed Google Scholar

  8. Daiana Barbosa da Silva
    View author publications

    You can also search for this author in PubMed Google Scholar

  9. Tarissa Ribeiro
    View author publications

    You can also search for this author in PubMed Google Scholar

  10. Cláudia Eugênio
    View author publications

    You can also search for this author in PubMed Google Scholar

  11. Daniel Schneider
    View author publications

    You can also search for this author in PubMed Google Scholar

  12. Débora Mariani
    View author publications

    You can also search for this author in PubMed Google Scholar

  13. Rodrigo Wiltgen Jeffman
    View author publications

    You can also search for this author in PubMed Google Scholar

  14. Fernando Bozza
    View author publications

    You can also search for this author in PubMed Google Scholar

  15. Alexandre Biasi Cavalcanti
    View author publications

    You can also search for this author in PubMed Google Scholar

  16. Luciano Cesar Pontes Azevedo
    View author publications

    You can also search for this author in PubMed Google Scholar

  17. Flávia Ribeiro Machado
    View author publications

    You can also search for this author in PubMed Google Scholar

  18. Jorge Ibrain Salluh
    View author publications

    You can also search for this author in PubMed Google Scholar

  19. José Augusto Santos Pellegrini
    View author publications

    You can also search for this author in PubMed Google Scholar

  20. Rafael Barberena Moraes
    View author publications

    You can also search for this author in PubMed Google Scholar

  21. Lucas Petri Damiani
    View author publications

    You can also search for this author in PubMed Google Scholar

  22. Nilton Brandão da Silva
    View author publications

    You can also search for this author in PubMed Google Scholar

  23. Maicon Falavigna
    View author publications

    You can also search for this author in PubMed Google Scholar

  24. Regis Goulart Rosa
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

DS, CT, CCR, RK, FB, ABC, LCPA, FRM, JIS, JASP, RBM, LPD, NB, MF, and RGR were involved in the conception and design of the SAP. DS, CT, and RGR drafted the manuscript. DS (fisrt author), CT, CCR, RK, MMSS, RMM, MGB, DBS, TSRH, CSE, DS, DMO, RWJ, FAB, ABC, LCPA, FRM, JIS, JASP, RBM, LPD, NBS, MF, and RGR revised and approved the final manuscript.

Corresponding author

Correspondence to Regis Goulart Rosa.

Ethics declarations

Ethics approval and consent to participate

This study will be conducted according to Resolution No. 466/12 of the Brazilian National Health Council (http://bvsms.saude.gov.br/bvs/saudelegis/cns/2013/res0466_12_12_2012.html). The present study protocol version (version 3, from 22 February 2017) has been approved by the research ethics committee of the coordinating site (approval No. CAAE 57717516.3.1001.5330) and the research ethics committees of all participating institutions. The need for patients’ written informed consent was waived in 37 of 40 participating ICUs, because the standard of care encompasses both study interventions. In 3 of 40 ICUs informed consent will be required for patients or proxies. Informed consent will be required for family members and ICU professionals in all ICUs.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Sganzerla, D., Teixeira, C., Robinson, C.C. et al. Statistical analysis plan for a cluster-randomized crossover trial comparing the effectiveness and safety of a flexible family visitation model for delirium prevention in adult intensive care units (the ICU Visits Study). Trials 19, 636 (2018). https://doi.org/10.1186/s13063-018-3006-8

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s13063-018-3006-8

Keywords

Trials

ISSN: 1745-6215