Effects of nurse-led cognitive behavioral therapy on insomnia in adults: a meta-analysis

Search of PubMed, Embase, Cochrane library, web of science databases for randomized controlled studies on the effect of nurses on insomnia was performed from the time of database creation to 25th of January 2025. Using the mesh word combined with a free word: nurse, insomnia, cognitive behavioral therapy. Detailed search strategies are provided in Supplementary Material Table S1. There were no language restrictions in this study. In addition, we considered the gray literature and manually searched conference abstract and dissertation databases.

The inclusion criteria for this study were: people aged > 18 years who met the diagnostic criteria for insomnia, the intervention was Nurse-led CBT-I, the control group was a blank control, and the primary outcome indicators were: Pittsburgh Sleep Quality Index (PSQI); Insomnia Severity Index (ISI); the quality of life (QOL), and secondary outcome indicators were: Sleep efficiency (SE); Total sleep time (TST), the type of study included in this randomized controlled study.

Conference abstracts, meta-analyses, systematic evaluations, animal experiments, full text unavailable, and case reports will be excluded.

Two authors rigorously screened the literature based on predetermined inclusion and exclusion criteria. In case of disagreement, both authors resolved the issue through negotiation or sought third-party advice and negotiated a consensus. Information extracted from the included studies included the following key details: authors, year, country, sample size, gender, mean age, co-morbidities, interventions and outcomes. For articles with data that could not be extracted, we first contacted the corresponding author to obtain the data before deciding whether to include it. For cases where there are incomplete data or key information not reported in the included studies, we first attempt to contact the original authors to obtain the relevant data. If supplementary information cannot be obtained, analysis will be conducted based on the available data. If the results of intentionality analysis (ITT) are provided in the research report, we will give priority to extracting this part of the data. If it is not clear whether the report is an ITT analysis, the main analysis results provided by the study shall be included.

To determine the quality of our results, we selected the Graded Recommendations Assessment Development and Evaluation (GRADE) system to evaluate the evidence [17] for methodological quality. We considered five factors that could reduce the quality of the evidence, including study limitations, inconsistent findings (judged based on statistical heterogeneity index I² > 50% or significant differences in effect sizes and confidence intervals across studies), inconclusive direct evidence, inaccurate or wide confidence intervals, and publication bias (assessed by funnel plot symmetry and Egger regression test). In addition, three factors that could reduce the quality of evidence were reviewed, namely effect size, possible confounding factors, and dose-effect relationships. A comprehensive description of the quality of evidence for each parameter data is provided (Supplementary Material Table S2).

Two researchers independently assessed the risk of bias as low, unclear or high using the Cochrane Collaboration’s tools [18]. If there was any disagreement, a third person was consulted to reach consensus. The assessment included seven areas: generation of randomized sequences (selective bias), allocation concealment (selective bias), blinding of implementers and participants (implementation bias), blinding of outcome assessors (observational bias), completeness of outcome data (follow-up bias), selective reporting of study results (reporting bias), and other potential sources of bias. Each included study was assessed individually against these criteria. If a study fully met al.l the criteria, the risk of bias for that study was ‘low,’ indicating that the study was of high quality and had a low overall risk of bias. If a study partially met the criteria, its quality was categorized as ‘unclear risk’, indicating a moderate likelihood of bias. If a study did not meet the criteria at all, it was categorized as ‘high risk’, indicating a high risk of bias and low quality of the study. To present the results of the risk of bias assessment in a comprehensive manner, we used RevMan 5.4 software to generate a Risk of Bias Summary and a Risk of Bias Graph, which showed the results of the assessment of each study in each area of bias and the overall percentage distribution of all studies in each area of bias, respectively.

The collected data were statistically analyzed using Stata 15.0 software (Stata Corp, College Station, TX, USA). Heterogeneity between included studies was assessed using I2 values or Q-statistics. I2 values of 0%, 25%, 50%, and 75% indicated no heterogeneity, low heterogeneity, moderate heterogeneity, and high heterogeneity, respectively. If the I2 value was equal to or greater than 50%, a sensitivity analysis was performed to explore potential sources of heterogeneity. If heterogeneity was less than 50 per cent, analyses were conducted using a fixed-effects model. Standardized mean difference (SMD) and 95% confidence interval (CI) were used for continuous variables and risk ratio (RR) and 95% confidence interval (CI) for dichotomous variables. In addition, random effects model and Egger’s test were used to assess publication bias.

As shown in Fig. 1, PubMed, Embase, Cochrane library, and web of science databases were searched to retrieve a total of 7636 relevant articles, 2134 duplicates were removed by removing them, 5487 articles were removed by reading the title and abstract, and 5 articles were removed by reading the full text, of which Did not report the outcomes of interest (n = 2), the full text is not available (n = 3) and finally 10 randomized controlled articles [19‐28] were included.

Ten articles involved 1537 patients, 3 articles [23, 24, 27] from USA, one article [21] combined with schizophrenia, one article [23] combined with lung cancer, one article [26] combined with Open Heart Surgery, one article [19] combined with cardiovascular disease, age range 20-78.92 years. The duration of the intervention ranged from 10 to 45 min; the specific basic characteristics are shown in Table 1.

Ten articles mention the generation of random sequences and are therefore evaluated as low risk, and three articles give a clear account of the blinding method used (participants were double-blind) and are therefore evaluated as low risk, and the results of the specific risk of bias are shown in Figs. 2 and 3.

In this study, funnel plot and Egger test were used to detect publication bias, funnel plot (Supplementary material figure S6-S10), Egger test PSQI (P = 0.246), QOL (0.063), SE (0.075),TST (0.065) results suggested that there was no publication bias and ISI (P = 0.036) suggested that there was publication in ISI bias is more likely.

To the best of our knowledge the study explores the effects of nurse-led CBT-I on insomnia in adults, and the results were nurse-led CBT-I was able to reduce PSQI scores, reduce ISI scores, improve QOL scores, improve SE, but has no significant effect on TST. These results provide strong evidence-based support for the use of Nurse-led CBT-I in clinical practice, particularly in relation to non-pharmacological interventions for people with insomnia.

Insomnia is a prevalent health problem worldwide, affecting the quality of life and mental health of many adults [29]. The PSQI and ISI are commonly used clinical tools to assess insomnia symptoms and sleep quality, and both scores reflect the degree of sleep disturbance in patients [30]. The results of this study showed that nurse-led CBT-I significantly reduced PSQI scores and ISI scores with a standardized mean difference (SMD) of -1.95 and − 3.34, respectively. This suggests that nurse-led CBT-I interventions have a significant clinical effect on improving insomnia symptoms. Nurse-led CBT-I improves patients’ sleep quality by helping them identify and change negative sleep cognitions and behavioral patterns [31]. Despite the statistical significance of the SMD of PSQI and ISI, their clinical significance needs to be explored. According to previous literature, the MCID for the PSQI is approximately 3 points [32] and the ISI is 3–8 points [32]. In this study, the SMDs were combined due to the different scales, but in terms of effect size strength, an SMD of more than − 0.8 is usually considered a “large effect”, so it can be hypothesized that the intervention is clinically significant. However, due to the heterogeneity of the scales and the differences in the study population, the clinical generalizability of the intervention should be interpreted with caution. The core components of nurse-led CBT-I include educating patients about the basics of sleep, developing sleep hygiene, relaxation training, and cognitive restructuring. These methods can help patients adjust their misperceptions about sleep, reduce anxiety before going to sleep, and improve the sleep environment, thus effectively reducing the severity of insomnia symptoms [33]. Compared with medication, nurse-led CBT-I has fewer side effects and longer-term efficacy, making it a recommended treatment for insomnia patients [34].

Improvement in quality of life is one of the core goals of treating any disease or condition. This study found that nurse-led CBT-I significantly improved patients’ quality of life [35]. This result suggests that nurse-led CBT-I interventions not only alleviate patients’ insomnia symptoms but also improve their overall quality of life. Insomnia patients are often accompanied by emotional problems such as anxiety and depression, and improved sleep quality helps to reduce these negative emotions and improve patients’ daily functioning and well-being [36]. In addition, nurse-led CBT-I may indirectly improve patients’ mental health status by reducing anxiety and improving sleep structure, thus contributing to quality of life [37]. SE is an important indicator of patients’ sleep quality, reflecting the efficiency of time allocation after falling asleep. The results of this study showed that nurse-led CBT-I significantly improved sleep efficiency, a finding that suggests that through cognitive behavioral therapy, patients can use their sleep time more efficiently and have less difficulty falling asleep and maintaining sleep, thereby improving overall sleep quality [38]. This finding is consistent with previous studies suggesting that nurse-led CBT-I has an important role in improving sleep efficiency in patients with insomnia [39]. Notably, some of the effect sizes (especially the ISI) were very large (SMD = -3.34) in the Meta-analysis of this study. Although this may reflect the significant efficacy of nurse-led CBT-I in improving insomnia, especially in specific populations, such large effect sizes also suggest that we need to interpret the results with caution. On the one hand, some of the literature included in this study had a small sample size, which may have a “small sample effect”; on the other hand, according to the asymmetry of the funnel plot and the results of the Egger’s test, there is a certain risk of publication bias in the ISI. Therefore, the combined effect value of ISI may be overestimated. In the future, it is still necessary to conduct a large-sample, well-designed randomized controlled trial to further verify the real effect of the intervention and reduce the risk of bias.

TST is a key indicator for assessing patients’ sleep status, reflecting the average number of hours of sleep per night. However, the results of this study suggest that nursing-led CBT-I failed to significantly improve TST. Although the change was relatively small, this does not mean that CBT-I had no effect on sleep duration. It is important to note that the main goal of CBT-I is to improve the quality of sleep (reduce nocturnal awakenings, improve time to fall asleep) rather than simply prolonging sleep duration, which may be one of the reasons for the lack of significant improvement in TST [40]. Indeed, patients with insomnia often face problems such as difficulty falling asleep or waking up during the night, and while CBT-I has shown promising results in addressing these problems, its impact on prolonging total sleep time is likely to be more limited [41]. Thus, the stability of the TST may reflect the fact that CBT-I is more focused on the goal of enhancing sleep quality rather than simply prolonging sleep duration.

CBT improves insomnia through a variety of mechanisms, the most important of which is cognitive restructuring. Patients with insomnia often hold negative sleep cognitions that exacerbate anxiety and disrupt sleep, and CBT helps patients identify and challenge these negative cognitions, prompting them to adopt positive ways of thinking that reduce anxiety and improve sleep quality. Nurses enhance patient acceptance of these cognitive changes through individualized coaching, further promoting efficacy [42]. Behavioral adjustment is also one of the key mechanisms of CBT. Insomnia patients usually have irregular sleep behaviors, and CBT helps patients restore regular sleep through behavioral interventions such as sleep hygiene education, stimulus control techniques, and relaxation training [43]. Emotional support from nurses helps patients adhere to these new behaviors, thereby improving outcomes. In addition to cognitive and behavioral mechanisms, emotion regulation plays an important role in CBT. Nurses help patients manage their emotions by building a trusting relationship, which promotes sleep quality by reducing the negative impact of mood swings on sleep [44]. Finally, CBT may improve insomnia through neurobiological mechanisms. Studies have shown that CBT alters brain activity, particularly in brain regions that regulate emotional and stress responses, such as the prefrontal cortex and amygdala, and promotes sleep quality by improving the balance of neurotransmitters (increasing levels of 5-hydroxytryptamine and norepinephrine) [45].

The results of this study suggest that nurse-led CBT-I is significantly effective in improving insomnia symptoms in adults. Given the wide-ranging impact of insomnia on adult health and the effectiveness of CBT-I, it is recommended that CBT-I be incorporated into nurses’ training. Nurse-led CBT-I interventions not only help patients to improve their sleep quality, but also provide individualized treatment, which in turn improves patient adherence and satisfaction with treatment. The benefit of implementing this intervention is that nurses are typically in a role closer to the patient on the healthcare team and can develop a better trusting relationship with the patient, which contributes to improved treatment outcomes. In addition, nurse-led CBT-I provides a viable and cost-effective treatment option that is particularly suitable for replication in resource-limited settings.