Study procedures
The study employs a modified Zelen design for consent [
17,
18]. Study staff will select, randomize, and intervene with patients prior to collecting verbal consent at the 1-month post-discharge visit. For the outcome analysis, the study uses a Bayesian adaptive trial design [
19]. Mediation analyses will also provide insight into why changing the default did, or did not, work. We hypothesize that smoker perceptions of what course of action their cessation counselor recommends, and what is the status quo for receiving medications and counseling at the hospital, will affect their counseling participation, medication use, and quit rates (Fig.
1). The trial will be implemented in four stages over 4.5 years (Table
1).
Table 1
Clinical trial timeline
6–23 | 1: Early implementation | • 400 participants consented • 6-month outcome data collected on 200 participants • Manuscript on study design under review |
24–39 | 2: Implementation | • Baseline data collected on 800 participants • 6-month data collected on 600 of participants • Data cleaned and prepared for analysis on half of all participants |
40–51 | 3: Early analyses | • Complete consent of final 200 participants by midyear • 6-month data collection completed on all participants • Manuscripts on baseline characteristics of participants published |
52–60 | 4: Analysis/dissemination | • All data cleaned and prepared for analysis • Conduct all outcome, mediator/moderator analyses • Manuscripts on study outcomes under review |
Recruitment of study participants
Random selection for the trial via the electronic health record (EHR)
Study staff have access via the EHR to a real-time, comprehensive list of all smokers in the hospital—the tobacco use list. For the purposes of the study, research staff will randomly select patients from the tobacco use list, and provide selected patients’ names to study staff for baseline assessment, randomization, and intervention. Random selection of participants serves two purposes. First, it ensures that we will test the change of default among a sample representative of all hospitalized smokers, which will enhance the generalizability of the findings. Second, it ensures that smokers not seeking tobacco-cessation treatment will be included in the trial, which will enhance our ability to detect the effect of changing the default among smokers who have not requested, and who might not be considered “ready” for medications and counseling.
Patients excluded from random selection via EHR
A number of hospital services routinely do not permit a majority of their patients to be placed on Nicotine Replacement Therapy (NRT). These include the operating room, orthopedic surgery, gastro-intestinal surgery, the burns service, and the neurology stroke service. Patients on these units, as well as patients less than 18 years of age, patients who have been in the hospital longer than 3 days, and pregnant patients, will be excluded from the tobacco use list prior to random selection for the trial. Patients who requested, or have orders for, tobacco-cessation treatment will be treated as usual by the existing hospital tobacco-cessation treatment service and will only be included in the trial if randomly selected.
Eligibility and baseline assessment
Study staff will visit all randomly selected smokers. For all study patients, staff will assess and address patient comfort, provide brief advice to quit, and provide a quit-smoking pamphlet. Staff will then assess eligibility and collect baseline data. Initial study eligibility criteria include: (1) the ability to speak English or Spanish, (2) have no significant comorbidity that precludes participation (i.e., acute, life-threatening illness or altered mental status such as dementia), (3) be a permanent resident of the state of Kansas or Missouri, (4) provide a secondary telephone contact to ensure 1-month follow-up survey completion, and (5) completed all eligibility questions.
Other eligibility criteria ensure that patients are able to benefit from the intervention. Because follow-up counseling is by telephone, participants are required to have access to a telephone or mobile phone. Because patients are receiving free starter packs of NRT and prescriptions for post-discharge smoking cessation medications, eligibility is limited to patients who smoke one or more cigarettes per day, have smoked at least 25 of the past 30 days, and are otherwise medically eligible to use NRT (i.e., no acute myocardial infarction/ST-elevation myocardial infarction (STEMI), or other acute heart conditions).
Last, several criteria preserve our ability to detect the effects of the trial. Patients cannot already have been seen by hospital tobacco-cessation treatment (UKanQuit) counselors or study staff during the recruitment hospitalization, cannot currently be taking cessation medications, or cannot be enrolled in a separate quit-smoking program.
Study staff will record eligibility for the study on their clinic service tablet computer, along with standard service administrative data, which includes demographics, smoking characteristics, readiness to quit, and contact information for the 1-month follow-up. These administrative data will constitute baseline data for the clinical trial.
The patients who are not eligible will be provided with the typical hospital tobacco-cessation services but will not be enrolled into the trial. The typical hospital services will follow the guidelines of UKanQuit [
20], which is a dedicated tobacco-cessation treatment service funded by the hospital. UKanQuit services include: (1) working with the medical team to address inpatient withdrawal symptoms, usually with NRT, (2) assessing readiness to quit, (3) providing motivational counseling to smokers who are not ready to quit, and (4) arranging for post-discharge counseling and medication prescriptions for patients who are ready to quit. The principal investigator (PI) of the present study also directs UKanQuit, and staff for both projects will work together closely to ensure that the study and the UKanQuit service avoid duplicating intervention among the same patients.
Enrollment and random allocation
All patients who are eligible will be automatically enrolled into the trial and randomly allocated to a treatment arm. Consent for the trial is delayed—it is sought at the 1-month follow-up. A function will be programmed into the tablet intake form so that study staff will select a key to randomize eligible smokers to either OPT OUT or OPT IN. At the beginning of the trial, the function will assign patients to groups in a 1:1 allocation ratio. Later, based on intermediate outcomes, the study biostatistician may alter the ratio in accordance with the Bayesian study design (see “
Data analysis,” below). Study staff will offer smokers motivational counseling, medications, and cessation-oriented practical counseling in accordance with the study arm to which patients are randomized. Patients are only randomized into our trial once. If someone has been randomized into the trial in the past—regardless of whether they provided consent at the 1-month follow-up or not—they are not eligible for repeat randomization. There are likely cases where a randomly selected patient is screened and found ineligible at one point in time, then eligible at a later readmission date. These patients may be enrolled at the later admission date.
Rationale for conducting a study with delayed consent
The study uses a modified Zelen design, in which consent is obtained after randomization and treatment. We considered consenting patients before randomization and treatment. This would, however, require patients to “opt in” to being in a trial and possibly exclude the very smokers who might benefit from changing the default. We rejected consent before randomization and treatment, as it is the very experimental manipulation we seek to test. The Zelen design, in which patients are consented after randomization, can markedly enhance recruitment and improve generalizability by including a higher proportion of eligible patients [
21,
22]. In a departure from most traditional Zelen design trials, the proposed study will delay consent until after treatment in order to examine the impact of OPT OUT on cessation at 1-month post discharge.
OPT IN and OPT OUT intervention procedures—framing the default
We have created language that constitutes the “choice architecture” for each study condition (Table
2)
. We have crafted these phrases to be short and simple, to enable study staff to reliably use them. Based on the patients’ group assignment, staff will frame the default and provide the appropriate intervention.
Table 2
Treatment options by study arm
Inpatient counseling and treatment plan—OPT OUT
For all patients in OPT OUT, staff will provide brief practical counseling, complete a treatment plan, and provide a pamphlet that outlines tips on quitting. The treatment plan includes: (1) inpatient medications, (2) outpatient prescription for medication and a free “starter pack” of medications, and (3) post-discharge counseling. In describing the treatment plan for the patients, study staff operationalize constructs thought to underpin the power of the default [
23]. These terms: (1) signal the provider’s positive attitude towards medication and counseling, and (2) state that the hospital’s status quo is to provide medications and counseling.
Using this OPT OUT language, staff will clearly indicate that they believe medications and counseling will benefit the patient, and that the hospital routinely and proactively provides cessation-oriented care to smokers. Unless a patient “opts out” of any or all elements of the treatment plan, all patients will receive all elements of the treatment plan, including inpatient medications, outpatient prescription and starter pack, and post-discharge counseling calls.
Inpatient counseling and treatment plan—OPT IN
For patients in OPT IN, staff will screen for willingness to stay having quit after leaving the hospital. Patients who are willing to quit will be offered similar counseling, treatment planning, and medications as patients in OPT OUT, including a pamphlet with tips on quitting. However, instead of proactively providing these services, staff will ask patients if they would like each element of the treatment plan—post-discharge counseling and/or post-discharge prescription for medication. Patients will only receive the elements of care to which they opt in.
Patients in OPT IN who are not willing to quit will receive brief intervention, in accordance with current treatment guidelines. This consists of a four-page pamphlet with resources for quitting and a brief counseling session that addresses the “5 Rs” of motivational counseling (
Relevance, Risks, Rewards, Roadblocks, Repetition) [
3].
Post-discharge treatment—OPT OUT—counseling
All patients in OPT OUT will be enrolled in post-discharge counseling that will be delivered by study staff. Participants who accept enrollment into counseling services will receive up to four proactive counseling calls in each of the 4 weeks following discharge. Each call is designed to provide practical counseling to help participants develop problem-solving and coping skills, secure social support, and design a plan for successful cessation and long-term abstinence. Initial calls last approximately 30 min and follow-up calls last on average 15 min. Once participants quit smoking, counselors review high-risk situations, coping skills, and stress management to prevent relapse. When participants “slip,” counselors troubleshoot relapse situations and encourage smokers to quit again. We will use RedCAP to store data for all calls, including number of attempts to reach the smoker, number of calls completed, and duration of calls.
Post-discharge treatment—OPT OUT—medication prescription and NRT “starter pack”
Study staff will work with all patients in OPT OUT to select a long-term cessation medication and plan how they will obtain and fill prescriptions post discharge. Study staff will request a prescription for post-discharge cessation medication via a note in the patients’ medical record, and via a text message to the floor pharmacist assigned to the patients’ medical team. The patients’ hospital physician will make the final determination regarding whether to write a cessation medication prescription, and which medication to prescribe.
To ensure that patients leave the hospital with some form of cessation medication, in order to avoid relapse, study staff will provide all OPT-OUT patients with a 2-week starter pack of over-the-counter quit-smoking medications. This will include 14 days of combination nicotine replacement pharmacotherapy, consisting of 14 nicotine patches plus 14-day supplies of either (a) nicotine gum, or (b) nicotine lozenges. The choice of the short-acting NRT will be made based on contraindications, past history of success/failure, and personal preferences [
3]. On the day of hospital discharge, study staff will provide the sealed starter pack to the patient at the bedside, for the patient to use once they leave the hospital.
Post-discharge treatment—OPT IN—counseling and medications
Study staff will provide OPT-IN patients who “opted in” to post-discharge counseling the same counseling that is provided to OPT-OUT patients. Likewise, for OPT-IN patients who want a prescription for post-discharge medications, study staff will arrange for a prescription and provide a starter pack in the same manner as these elements are provided to OPT-OUT patients.
Month-1 call for service data collection and informed consent
In accordance with standard procedures in the existing UKanQuit treatment program [
20,
24], study staff will call all patients at 1 month post enrollment to assess outcomes including smoking status, quit attempts, counseling utilization, medication use, and other factors related to quitting (see Table
3, Core study measures). At the close of the call, study staff will verbally debrief patients on their inclusion in the trial and invite them to participate in the study. Study staff will collect additional data related to study outcomes.
Table 3
Core study measures
Outcomes |
7-day point-prevalence abstinence | | ✓ | ✓ |
Biochemical quit verification | | ✓ | ✓ |
Number of quit attempts since enrollment | | ✓ | ✓ |
Post-discharge counseling adherence/other support | | ✓ | ✓ |
Medication use/adherence | | ✓ | ✓ |
Sociodemographics/mediators/moderators |
Demographics: age, gender, race | ✓ | | |
Length of hospital stay (for index visit) | ✓ | | |
Reason for hospitalization (index visit) | ✓ | | |
Readiness to quit, craving/withdrawal | ✓ | ✓ | |
Number of cigarettes per day (cpd); time to first cigarette | ✓ | ✓ | |
Motivation/confidence quit/stay having quit | ✓ | ✓ | |
Default constructs (perceived status quo, etc.) | | ✓ | |
Rehospitalization | | ✓ | ✓ |
Cost measures |
Counseling | | ✓ | ✓ |
Nicotine Replacement Therapy (NRT) (calculated from patient self-reported use) | | ✓ | ✓ |
Rationale for measuring main outcome at 1 month post intervention
The present study is focused on how best to engage smokers in tobacco-cessation treatment (i.e., counseling and medications) and quit. Assessing outcomes at 1 month will best capture the immediate impact of OPT IN versus OPT OUT on medication/counseling use, quit attempts, and abstinence [
25,
26]. Moreover, assessing outcomes at 1 month is in accordance with Joint Commission Guidelines and UKanQuit standard practice for post-discharge follow-up of hospitalized smokers [
27]. Perhaps most importantly, 1 month represents an important timeframe for hospitals, as hospitals with excessive 1-month readmission rates for selected diagnoses will receive decreased Medicare reimbursements [
28]. Should our intervention prove effective, it could pave the way for future studies on the impact of smoking cessation on reduced 30-day readmission rates for specific diagnoses. Those who refuse consent at 1 month will not be eligible for inclusion again into the trial and they will be included (simply counted as smokers, because we will have no data on them) in our simplified intent-to-treat (ITT) analyses.
Reimbursement
Patients who participate and complete the 1-month survey will be reimbursed US$25 whether or not they consent to participate in the clinical trial. Study staff will reimburse patients who consent to participating in the clinical trial and completing the extended 1-month survey with an additional US$25. All reimbursements will be via reloadable debit cards. The debit cards utilize the MasterCard payment system and are accepted at virtually every institution that accepts a credit card. Participants will also be reimbursed US$25 for the 6-month follow-up survey and US$50 for each salivary cotinine sample returned.
Project measures (Table 3)
Tobacco abstinence
Outcome measures are adapted from the Society for Research on Nicotine and Tobacco’s Workgroup on Abstinence Measures and Workgroup on Biochemical Verification [
29,
30]. Our primary endpoint is 7-day, verified cigarette abstinence at 1 month after enrollment. Subjects with missing data will be counted as smokers.
Verification of abstinence
We will use either mailed salivary cotinine or in-person carbon monoxide (CO) testing to confirm smoking status. Participants who report 7-day point-prevalence abstinence, and who are not taking NRT, will be asked to provide a saliva sample. Cotinine is the measure of choice because of its sensitivity and specificity [
30]. Samples will be stored in a − 20 °C freezer until laboratory analysis. Participants who are still using NRT, or who refuse salivary cotinine, will be verified via CO testing. Those with
< 10 ppm will be considered abstinent.
Study counselors will track counseling data, which will be summarized as “total counseling time” for analyses. We will assess the type, the dose, and the number of days of medication was used via the method of Williams et al. [
31,
32]. This will be summarized as “number of days of medication use” for analyses. Default-related variables are derived from the academic literature on choice theory and include smokers’ perceptions of provider attitudes toward tobacco-cessation treatment (implied recommendation), smokers perceptions of the degree to which their provider recommends tobacco-cessation treatment (implied recommendation), perceptions of the “status quo” for hospital tobacco-cessation treatment (status quo bias), and perceptions of paternalistic treatment by UKanQuit staff [
13,
23,
33].
Intervention costs
We will prospectively track variable intervention costs. Costs will include inpatient counselor services, post-discharge counseling time, and initial pharmacotherapy dispensed at baseline. During the 6-month follow-up call, we will ask participants to recall their use of pharmacotherapy after the initial supply. Personnel time will be valued at Bureau of Labor Statistics (
www.bls.gov) wages plus benefits for an appropriately trained health-promotion professional. Pharmacotherapy costs will be based upon retail prices estimated from on-line pharmacy websites. Intervention costs will be tracked as they are incurred. We will exclude research costs. We will not discount either costs or benefits.
Fidelity monitoring
Fidelity to the study protocol will be assessed by in-person fidelity assessments during hospital consults. To assess the quality of the intervention and control conditions, we will assess the degree to which study staff accurately: (1) provide brief advice and (2) perform the appropriate intervention for OPT OUT versus OPT IN. Data on fidelity will be entered into a RedCAP database and reported back to study staff to encourage adherence to protocols.
Protection against risks
There are minimal risks in this study. Any emails used to transmit study participant information will be encrypted to protect the privacy of patients. Salivary cotinine samples are noninvasive; samples will be labeled with participants’ study ID numbers, rather than names, to protect participants’ privacy. When collecting proxy verification of smoking status, no information about the patients’ participation in the trial will be provided to the proxy besides the participants’ name and the fact that they had nominated the proxy to provide verification of smoking status at the end of a health-promotion study.
Standard language in our consent procedure assures the participants of the confidential nature of the study. Those who elect to participate will be clearly told that they may withdraw from the study at any time without jeopardizing current or future care at any medical facility. Potential participants will also be informed of alternative treatments (i.e., using other smoking-cessation programs, purchasing nicotine gum, patches, or lozenges from the pharmacy, obtaining a prescription for a nicotine inhaler, nasal spray, or other smoking-cessation products from their physician). These standards are strictly adhered to and monitored by the KUMC Institutional Review Board. Only summaries of group data will be reported in any publications or presentations, with no identification of individuals. All records will be kept in locked filing cabinets in offices that are kept locked when unoccupied. Subject files will be kept in a secure area, with access only by designated staff members (PI and co-investigators).
Data management
Data management will follow procedures developed for Enhancing Quitline Utilization among In-Patients (EQUIP). UKanQuit service data, and survey data collected by research assistants, will be directly entered via tablet into REDCap. The project director will coordinate data retrieval from the EHR. The data manager will conduct initial data cleaning, identifying and tagging any crossovers, conversion into proper format for data analysis, and recoding using standard operating procedures. All data will be imported into SAS for study analyses. Cleaning and management routines (e.g., conversion of birth dates to ages, logical checks for continuous variables, compliance with skip patterns, missing data codes) will be conducted using SAS.
Data safety and monitoring plan
Due to the low level of risks involved in the proposed study, a Data and Safety Monitoring Board (DSMB) will not be necessary. The data and safety monitoring will be overseen by the PI, Dr. Richter, and an annual progress report will be provided to the Human Subjects Committee of the University of Kansas Medical Center as well as the NIH. The purpose of this data and safety monitoring plan is to ensure the safety of study participants and the validity of data in compliance with National Institutes of Health (NIH) requirement of Data and Safety Monitoring for Clinical Trials. This section outlines essential elements of the Data Safety and Monitoring (DSM) plan for this clinical trial.
Plans for assuring data accuracy and protocol compliance
Data management activities for this project will encompass data entry, data cleaning, identifying and tagging any crossovers, conversion into proper format for data analysis and recoding. In addition, a computer-based tracking system will be developed to follow each patient and to prompt the staff for the upcoming data collection point. Data collection points for each subject will be calculated from their initial date of contact. Data entry will be performed on site under direction of the PI and the study statistician. Keypunching routines will adhere to the codebook specifications written by the team. Codebooks will include variable formats (numeric/alpha), min/max ranges and any skip patterns. Data will be double entered utilizing two separate databases and data entry personnel for the two databases would be different. Both the databases and the tracking system would be password protected for security and maintenance of confidentiality. At the end of each data entry period, data would be backed up onto a storage unit. Checks built into the database will ensure that individuals not meeting eligibility criteria are flagged and excluded from data analysis.
Data sharing
This study will generate quantitative and qualitative data from the randomized trial. The final quantitative dataset will include self-reported demographic and behavioral data from subjects. Because we collect data at multiple time points from participants over the 6-month period that each participant will be in the study, we will collect identifying information. Prior to data sharing, we will remove or convert all identifying information (date of birth will be converted to age, date of admission and discharge will be removed, and other identifiers will be removed). There may remain the possibility of deductive disclosure of subjects with unusual characteristics. Thus, we will make the data and associated documentation available to users under a data-sharing agreement that provides for: (1) a commitment to using the data only for research purposes and not to identify any individual participant, (2) a commitment to securing the data using appropriate computer technology, and (3) a commitment to destroying or returning the data after analyses are completed. Data will be saved as SAS or SPSS files, saved to disk, and mailed to users.