Discussion

Our findings revealed that the prescribing rate of guideline-recommended pharmacotherapy was low for patients with insomnia comorbid depressive or anxiety disorders at primary healthcare facilities in China. Benzodiazepines and non-benzodiazepines were the most commonly prescribed medications, despite clinical guidelines prioritising antidepressants with sedative effects, SSRIs and SNRIs, which were less frequently used. ⁷

The prescribing rates of psychotropic medications for patients with insomnia comorbid with depression or anxiety exceeded 90% overall and across different age groups. Psychotherapy, such as CBT, has been demonstrated to be effective for common mental disorders (e.g. insomnia, depression, anxiety, solely or comorbid), and is recommended as first-line treatment in current clinical guidelines. ^{Reference Qaseem, Owens, Etxeandia-Ikobaltzeta, Tufte, Cross and Wilt9,Reference Bandelow, Allgulander, Baldwin, da Conceição Costa, Denys and Dilbaz10,Reference Qaseem, Kansagara, Forciea, Cooke and Denberg25,Reference Riemann, Baglioni, Bassetti, Bjorvatn, Dolenc Groselj and Ellis26,} However, access to psychotherapy remains extremely limited worldwide particularly in primary healthcare facilities, because of inadequate resources and a shortage of trained therapists. ^{Reference Kyle, Siriwardena, Espie, Yang, Petrou and Ogburn11,Reference Kendrick12} Given these constraints, patients with mental disorders were frequently treated only pharmacologically. ^{Reference Kendrick12,Reference Fu, Zhu, Gong, Li, Li and Shi14} Enhancing access to psychotherapy may help reduce unnecessary psychotropic medication use and mitigate associated adverse effects and drug-drug interactions, which is particularly pertinent to older persons. ^{Reference Nimmons, Bazo-Alvarez, Avgerinou, Hayes, Osborn and Cooper22}

Our findings highlighted that only 29.9% of patients with comorbid insomnia and depression and 11.5% of those with comorbid anxiety received treatments aligned with guideline recommendations. We observed high overall use of psychotropic medications across all age groups, but guideline-recommended prescribing rates were lower among adults aged ≥65 years, particularly among those with insomnia and depressive disorder. In older populations, sedative hypnotics should generally be avoided because they are associated with falls, delirium, cognitive impairment and dependence. ²⁷ Given the high prevalence of multimorbidity and polypharmacy in older adults, primary care should routinely review medications and screen for cumulative sedative and anticholinergic burden. Expanding access to CBT for insomnia and, when syndromic depressive or anxiety disorders are present, considering antidepressants with sleep-promoting properties may reduce reliance on chronic hypnotics in this high-risk group.

Another major concern is the high prescribing rates of non-benzodiazepines and benzodiazepines, both of which should be used with caution according to the Chinese guideline. ⁷ However, our results showed that non-benzodiazepines were prescribed for 42.5% of patients with insomnia comorbid depression and 42.4% of those with comorbid anxiety, representing substantial inappropriate use. Moreover, although benzodiazepines were considered a last resort by guidelines because of their high risk of inducing cognitive dysfunction and falls, ^{Reference Billioti de Gage, Moride, Ducruet, Kurth, Verdoux and Tournier28–Reference Maust, Lin, Goldstick, Haffajee, Brownlee and Bohnert30} they were prescribed at alarmingly high rates in both conditions (55.9 and 69.6%), far exceeding rates reported in the UK (14.5–27.1%). ^{Reference Nimmons, Bazo-Alvarez, Avgerinou, Hayes, Osborn and Cooper22} This pattern of high non-benzodiazepine and benzodiazepine use, particularly noted in patients aged >85 years, has also been reported in our previous research focusing on insomnia treatment in primary care. ^{Reference Fu, Zhu, Gong, Li, Li and Shi14} This pervasive issue may stem from clinicians’ limited knowledge of guideline recommendations in managing comorbid common mental disorders. ^{Reference Fu, Zhu, Gong, Li, Li and Shi14} Tailored training programmes are therefore necessary to promote evidence-based treatment at primary healthcare facilities and reduce excessive use of benzodiazepines, especially among older adults.

The prevalent use of benzodiazepines or non-benzodiazepines as monotherapy in treating comorbid conditions suggests an oversight of depressive and anxiety symptoms by primary clinicians. This issue may primarily arise from primary care physicians’ unfamiliarity with the guidelines as discussed earlier, but it might also be influenced by patients’ expectations. ^{Reference Sirdifield, Anthierens, Creupelandt, Chipchase, Christiaens and Siriwardena31} Some patients who initiated appropriate treatment at specialty hospitals or tertiary hospitals may seek stronger sedative hypnotics at primary care. This highlights the need for primary clinicians to be trained to meticulously review patients’ psychotropic medication use during consultations and make necessary medication reconciliation to ensure more appropriate prescribing and patient safety. ^{Reference Almanasreh, Moles and Chen32} Establishing a shared network system among different healthcare levels to track diagnosis and treatment history could facilitate more comprehensive care. ^{Reference Tian, Chen and Bai33} Future studies using regional databases that link different levels of healthcare institutions are need to better understand cross-setting differences in pharmacotherapy for insomnia comorbid with depression or anxiety.

Our research also showed prevalent use of flupentixol-melitracen in primary healthcare facilities in China, a pattern not observed in other countries. ^{Reference Nimmons, Bazo-Alvarez, Avgerinou, Hayes, Osborn and Cooper22,Reference Tanana, Harrison, Nishtala and Chen34} Flupentixol-melitracen is a fixed-dose combination of a typical antipsychotic and melitracen, a tricyclic antidepressant. It has not been marketed in many high-income countries (e.g. USA, Canada, Japan, Denmark), and was banned by India’s health ministry in 2015 because of safety concerns regarding fixed-dose combinations of psychotropic agents. ^{Reference Bhaumik35} Despite these international reservations, flupentixol-melitracen continues to be widely used in China, where it has been marketed since the late 1990s. Although Chinese guidelines recognise its efficacy for treating depression and anxiety, they offer limited guidance on its use in managing insomnia comorbid with depression or anxiety. The substantial use of this combination in the absence of definitive and conclusive evidence raises concerns about potential harms. Rigorous clinical research is urgently needed to establish clear evidence to inform and refine future treatment recommendations. Further, Chinese guidelines should provide more specific and concrete recommendations for flupentixol-melitracen to support safer clinical practice and improve patient outcomes.

Several limitations of this study should be noted. First, the data sources contained only primary healthcare facilities, whereas the treatment of depression and anxiety might predominantly occur in specialised or tertiary hospitals in China. This may explain the relatively small sample size of patients with insomnia comorbid depression or anxiety in our study, limiting its generalisability. Future region-wide, cross-setting linked studies are needed to assess duration and trajectories of care. Second, potential inaccuracy or inconsistency of diagnostic coding in the database may have affected case identification and inclusion. However, we also used the corresponding Chinese text diagnoses recorded in the electronic health records, which helped to partially mitigate this uncertainty in coding. Third, the study was conducted in Dongcheng, a district in central Beijing with higher economic status and superior medical resources. Consequently, the findings may not be fully representative of the broader population of patients in primary care across Beijing. Fourth, we restricted our study period to 2022 to maximise data completeness and coding stability. A single-year analysis limits assessment of trends such as seasonality or secular shifts.

In conclusion, the prescribing rate of guideline-recommended pharmacotherapy for patients with insomnia comorbid with depressive or anxiety disorders was considerably low in Chinese primary healthcare facilities, primarily because of the excessive use of benzodiazepines and non-benzodiazepines. These findings highlight the need to reduce the use of high-risk medications and to facilitate evidence-based treatment for complex mental health conditions at primary care, especially for older adults.