In contrast, longitudinal analysis models the daily quit probability, rendering the interpretation independent of other elements of the study design.
As the longitudinal analysis uses all time points in the treatment period, the drug effect OR represents an effect averaged across time, whereas the standard analysis reflects the effect only at EOT. Our data analysis revealed higher ORs with the longitudinal models, suggesting some variation of drug effect over time. Regardless of the estimated ORs, the corresponding CIs from longitudinal models are generally narrower than those from simple logistic models, with the size of the difference depending on the within-subject correlation [ 20 ].
A longitudinal analysis of daily smoking status has greater potential efficiency gain when some observations are missing, because it can include the available data from all randomized subjects, even those lost to follow-up, whereas the standard approach has to either exclude the drop-outs or assume that they continued to smoke. Although such an assumption is held to be conservative, there is an increasing awareness that its indiscriminate use may lead to bias and lack of comparability between studies [ 26 , 27 26]. Fortunately our example had few missing observations. An important advantage of longitudinal modeling is its ability to incorporate time-varying predictors.
In some studies, treatment such as drug dose changes over time by design [ 28 ]. Even if the treatment is constant, including the treatment-by-time interaction allows us to test whether its effects change over time, as in our example. We investigated the influence of smoking history on later success by coding summary measures of history as time-varying predictors. Our results suggest that history is an important independent predictor of future outcome. Generally, there is concern that including outcome history may over-adjust and thereby attenuate treatment main effects.
In our results, the drug effect OR in the ME model was 2. Possibly the history variable absorbed some of the treatment effect, resulting in the decreased OR. Still, the large size and strong statistical significance of the adjusted effect suggests that bupropion continues to have an effect no matter how long one has been taking it and regardless of its effect to date. Although the ME and GEE models work similarly in many cases, one must bear in mind that their regression coefficients have different interpretations [ 17 ].
Statistical Analysis of Daily Smoking Status in Smoking Cessation Clinical Trials
The drug effect OR in the ME model represents the odds of the outcome for a person taking the drug compared to the same person not taking the drug. Estimates from GEE are generally smaller than those from ME, and the attenuation increases with the between-subjects variability. After applying the scale factor as we illustrated, the two estimates are comparable [ 18 , 19 ].
A marginal approach like GEE is problematic when one wishes to model time-varying effects [ 21 , 22 ]. As shown in our example, GEE is unreliable when one includes the outcome histories as predictors. Moreover, ME is preferable when there is substantial dropout, because one can estimate it consistently with weaker assumptions on the missing-data mechanism [ 29 ].
Comparison of Physical Fitness among Smoker and Non-Smoker Men
A potential limitation of our analysis is that we used treatment phase data only. Given the logarithmic nature of relapse curves [ 30 ], outcomes at later follow-ups such as 6 or 12 months are considered superior indicators of treatment success. Commonly, however, daily smoking data are either not collected or are unreliable after EOT, limiting the use of daily data beyond that point.
Nevertheless, the analysis we presented provides a way to evaluate the dynamic process in the treatment period using all the available information; one may conduct separate analyses on later outcomes using standard methods. Another limitation of our analysis is its use of TLFB data, which are self-reported and thus may be inaccurate in a fraction of subjects who falsely report abstinence [ 31 ]. Nevertheless, our data suggest that the drug effect is similar whether we use self-reported or verified EOT point prevalence abstinence as the outcome Table 2.
One possible reason is that subjects in both arms are equally likely to report false abstinence and therefore any biases are offsetting. Models for daily abstinence would not be biased by subjects who under-report cigarette counts, as long as they do not claim abstinence on smoking days. TLFB data are typically collected every one or two weeks and are therefore potentially subject to recall bias, which may affect estimates of within-subject correlation if the subjects incorrectly report the same or very similar counts for all the days being recorded at each visit.
We observed the same large correlation, suggesting that it is real and not simply an artifact of recall bias. Moreover, the drug effect OR was similar to that from the daily data, with a slightly wider CI. This is expected because the efficiency gain from using daily vs. The advent of electronic diaries, allowing collection of cigarette counts by ecological momentary assessment, may obviate the need for daily summarization of cigarette counts.
One could in principle obtain a more efficient analysis from the series of daily cigarette counts. Because such data are subject to severe heaping, in the form of over-reporting of round numbers of cigarettes smoked [ 5 ], there is substantial potential for bias in such analyses, and it has been considered preferable to use the daily abstinence indicators. As we have suggested, another approach is to model the duration of abstinent and smoking episodes, incorporating the possibility of permanent recovery and relapse [ 6 ]. Such models, when estimated exclusively from treatment-period data, can make excellent predictions for remote long-term outcomes [ 32 ].
With current types of data, the day is still the smallest time interval, but as electronic recording devices become more common, it will be possible to measure inter-cigarette intervals to the second, providing for an even finer analysis. We have presented a strategy to model longitudinal daily smoking status data from smoking cessation trials. Compared to the standard analysis of EOT abstinence status, our approach permits more detailed modeling and more precise estimation of effects.
In our example study a strong drug effect persisted in the treatment period, with its magnitude varying across time. The National Cancer Institute supported the research of Drs. We are grateful to Dr. Lerman for permission to use the data, to the referees for constructive suggestions, and to Dr. Thomas Ten Have for enlightening discussions.
Each subject can have up to 56 rows of data. All authors declare that they have no conflict of interest. Clinical trial registration details: National Center for Biotechnology Information , U. Author manuscript; available in PMC Nov 1. Yimei Li , 1 E. Paul Wileyto , 2 and Daniel F. Author information Copyright and License information Disclaimer. The publisher's final edited version of this article is available at Addiction.
See other articles in PMC that cite the published article. Abstract Aims Smoking cessation trials generally record information on daily smoking behavior, but base analyses on measures of smoking status at the end of treatment EOT.
- Tolerance of Non-Smokers to Smokers.
- Comparison of Physical Fitness among Smoker and Non-Smoker Men.
- The Good Confession!
Methods We analyzed daily abstinence data from a smoking cessation trial, using two longitudinal logistic regression methods: Results We observed some differences in the estimated treatment effect odds ratios across models, with narrower confidence intervals under the longitudinal models.
Conclusion When analysing outcomes of studies from smoking cessation interventions, longitudinal models with multiple outcome data points, rather than just end of treatment, can makes efficient use of the data and incorporate time-varying covariates. Generalized estimating equations, longitudinal analysis, mixed-effects model. Table 1 Subject characteristics by treatment arm. Open in a separate window. It is based on t test for the continuous variable fraction of smoking days, and on the chi-squared test for all other variables, which are binary.
Longitudinal analyses with time-varying predictors We also conducted analyses including the time-varying predictors daily and weekly smoking history, a time effect, and a drug-by-time interaction. Footnotes Declaration of interest: A comparison of four self-report smoking cessation outcome measures. Measures of abstinence in clinical trials: Fifty ways to leave the wagon. Relapse and relapse prevention. Annu Rev Clin Psychol. Wang H, Heitjan DF. Modeling heaping in self-reported cigarette counts. Modeling smoking cessation data with alternating states and a cure fraction using frailty models.
A self-report depression scale for research in the general population. Bupropion for smoking cessation: Predictors of successful outcome. Bupropion SR and counseling for smoking cessation in actual practice: Nicotine Tob Res Gender differences in smoking cessation in a placebo-controlled trial of bupropion with behavioral counseling. Do small lapses predict relapse to smoking behavior under bupropion treatment? A mixed-effects model for categorical data. Longitudinal data analysis using generalized estimating equations. Mediating mechanisms for the impact of bupropion in smoking cessation treatment.
Recurrent event analysis of lapse and recovery in a smoking cessation clinical trial using bupropion. Carriere I, Bouyer J. Choosing marginal or random-effects models for longitudinal binary responses: Molenberghs G, Verbeke G.
Models for Discrete Longitudinal Data. Comparison of population-averaged and subject-specific approaches for analyzing repeated binary outcomes. Design effects for binary regression models fitted to dependent data. Lindsey JK, Lambert P. On the appropriateness of marginal models for repeated measurements in clinical trials. Furthermore, no communication with the outside world is permitted, which poses major psychological stress.
This situation offers a particularly interesting area of clinical study. Submarines are one of the only places where subjects are completely cutoff from the outside world for several months at a time. Submarine physicians general medicine specialists trained in remote medicine have noticed that submariner smokers who abruptly stop their tobacco addiction at the beginning of a mission do not suffer from withdrawal symptoms. However, they have found that submariners quickly resume smoking after their operational patrol has ended.
This paradox has not yet been scientifically investigated.
Studies into smoking in confined environments are rare, as they normally investigate passive smoking. This original study seeks to quantitatively measure withdrawal symptoms in smokers during a submarine patrol and to clarify the factors that promote smoking resumption upon their return to the surface. Between November and February , the all-male crews of two submarines departing on a mission of approximately 70 days were offered the opportunity to participate in the study. Inclusion criteria were their continued presence on board during the selected mission and their voluntary participation. Noninclusion criterion was the refusal to participate in the study.
The study consisted of five questionnaires.
The first questionnaire was administered during the medical examination before the patrol. The second questionnaire was completed 48 hours after beginning the mission, the third questionnaire was completed in the middle of the patrol, and the fourth questionnaire was completed 1 week before completing the patrol. The evaluation scores chosen in this study are validated scores. The Mood and Physical Symptoms Scale has proven to provide linkages between each symptom and smoking cessation. In the literature, three scales emerge: Their Cronbach indices are satisfactory 0. The MNWS was chosen because of its power and reliability, and it can be administered more quickly.
A score that progressively decreases during smoking cessation indicates the presence of withdrawal symptoms. In measuring the withdrawal score, it seemed of paramount importance not to forget social influences e. This score should not be interpreted in absolute terms, as it was used in situations for which it had not been validated, but rather as a measure investigating whether smokers who were in withdrawal were more anxious and depressed than nonsmokers.
Finally, only the smokers were invited to complete the fifth questionnaire during the 2 months following their return. This final questionnaire sought to evaluate the importance of smoking resumption after the patrol and reasons for smoking cessation failure. A maximum of 6 months elapsed between inclusion in the study and the final questionnaire. A verbal explanation of the study was first given to the entire cohort and was then given individually. Each study participant signed a written consent form in the presence of a physician investigator. Data collection was conducted in compliance with the Helsinki Declaration by assigning anonymous numbers and not collecting the participants' names.
Mean comparisons quantitative variables were conducted using the nonparametric Mann—Whitney and Kruskal—Wallis tests. Percent comparisons qualitative variables were conducted using the chi-squared and Fisher's exact tests. Statistical analyses were performed using R software http: All statistical tests were two-sided. In total, members of the crew were offered participation in the study and subjects agreed to participate; 52 The population was divided into a smoking cohort and a nonsmoking cohort nonsmokers and ex-smokers.
All subjects completed the first four questionnaires, whereas only 36 of the 52 smokers A comparison of demographic factors between the two cohorts is provided in Table I. The smoking population smoked an average number of cigarettes Within the smoking cohort, There was no significant difference in the HAD score between the smoking and nonsmoking cohorts in the different questionnaires.
Within the same population, there was also no significant variation in the HAD score among the different questionnaires.
- Personality and Smoking Behaviour of Non-Smokers, Previous Smokers, and Habitual Smokers.
- Statistical Analysis of Daily Smoking Status in Smoking Cessation Clinical Trials.
- Post your comment?
- Encounter with Tiber.
- Solar Power Practical Considerations (The Science of Electricity).
Analysis by age and rank subgroups did not offer any statistically relevant information. For the smoking cohort in general, the MNWS was stable over time. There was no significant variation in weight before and after the mission. During the completion of the questionnaires before and after the operational patrol, the number of smokers wishing to take up smoking again after discharge increased.
Among the 36 smokers who responded to the final questionnaire, 23 Therefore, there was a Those who resumed smoking and those who quit had the same mean age 29 years and the same distribution in rank. To one of the closed questions in the fifth questionnaire, 12 of the smokers The latter mentioned in an open question, the presence of smokers in their environment and the feeling of freedom as one of the reasons for resuming smoking.
Other commonly cited reasons for smoking resumption are inactivity upon return, absence of motivation, and a readily available supply of tobacco. An analysis of the kinetics of smoking resumption indicates that The attrition bias in the last questionnaire is important yet independent of resumption status.
The bias is due to transfers aboard other vessels at the end of the mission. Demographic data and smoking habits gathered from the first questionnaire and from those who responded to the fifth questionnaire as well as from other follow-up of smokers cannot be analyzed because of a low number of respondents.
In our study, there were no withdrawal symptoms in smokers during abrupt and forced smoking cessation. After 11 weeks of cessation, there is no longer a physical or psychological dependence on nicotine, indicating that the smoking resumption observed is solely attributable to a behavioral dependence. However, in literature, withdrawal symptoms usually became apparent because of the greater presence of signs of anxiety or depression in the smoking cohort, particularly during the first week following cessation.
Moreover, in the presence of withdrawal symptoms, the MNWS generally decreases along with smoking cessation. However, this was not the case in this study. Several explanatory hypotheses are included below: In our study, and according to the doctors survey, However, the comparison of biographical elements and smoking habits is not feasible due to a lack of participants.
As a result, we cannot conclude that the use of patches or not is efficient … even though we know that according to the relevant literature it is proven so. Finally, according to the investigative practitioners in the two submarines, the subjective feeling of the participants using patches was very good, without any noted undesirable incident. During the very serious medical expertise appointments before the missions, all the smokers are systematically offered patches.
It is worth noting that according to the first questionnaire, The patients may, therefore, have a feeling of efficiency from the patch in the prevention of withdrawal symptoms, but not in terms of efficiency of smoking cessation … although relevant data describes patches as efficient in both cases. The beginning of a mission, when withdrawal symptoms are strongest, is often a period of intense professional activity for the submarine crew.
It is by nature tiring and stressful, which likely obscures some of the sleep disorders observed during nicotine withdrawal.