Insomnia, defined as difficulty falling asleep, staying asleep, or waking up too early for 3 or more nights per week for 3 months or more with significant daytime effects,1 is the most common sleep problem presenting in general practice: 10 to 12% of the population meet formal diagnostic criteria for insomnia disorder, and up to 40% express complaints of insomnia.2
For those suffering from sleep problems, the GP is often the first point of contact, with previous research suggesting 79% of GPs see someone with a sleep complaint at least once a week.3 GPs are often unaware of patients’ sleep problems, and, even when they are, fewer than 8% of doctors use validated sleep questionnaires or sleep diaries.4 Despite both GPs and patients believing that a detailed sleep assessment is important, this lack of assessment may be due to time pressure and/or a limited knowledge of how best to evaluate sleep complaints.5
For these reasons, a brief measure that can reliably screen for insomnia could be invaluable. The two-item Sleep Condition Indicator (SCI-02) has been developed to help GPs and primary care nurses screen for insomnia6 (Table 1). The two items derived from the full, validated eight-item Sleep Condition Indicator (SCI)6 include questions reflecting being troubled about sleep problems and the frequency of the sleep complaint, and have been suggested for a short version based on their high predicted value (82% variance) of the full-scale SCI.6
Table 1. Two-item version of the Sleep Condition Indicator (SCI-02)
Each item is scored on a 5-point scale (0–4), with lower scores, in the 0–2 range, reflecting DSM-5 threshold criteria for insomnia disorder. Possible total scores range from 0–8, with higher values indicative of better sleep.
We have validated the SCI-02 using a sample of 190 000 persons who completed it online, randomly extracted from an online platform or mobile app (www.sleepio.com), similar to a previous validation of the full SCI.7 By completing the measures online participants agreed that their data could be used anonymously for research. A subsample of participants also completed the remaining questions of the full SCI within 1 hour, allowing us to assess the correlation between the SCI-02 and the full SCI.
The sample of 190 000 adults had a mean age of 40.24 ± 14.31 years and comprised 105 839 women (55.7%). Cronbach’s α and the Spearman–Brown correlation for the entire sample were both acceptable at 0.74. The test–retest reliability and intraclass correlation coefficient in a sample repeating the test from 12 hours up to 7 days were r = 0.68 and ICC = 0.68 respectively.
In a subsample of 4612 users (age: 41 ± 12; 57% female) who completed both the SCI-02 and the remaining six items of the full SCI within 1 hour, the SCI-02 was correlated strongly (r = 0.80) with the total score of the full SCI. A cut-off of ≤2 for the SCI-02 predicted those identified with probable insomnia according to the full SCI, with a specificity of 81% and sensitivity of 80%. Some caution is needed as the sample was self-selected, thus likely to be in favour of those who had an interest in sleep, and a bias towards those with a sleep problem.
The ultra-short, two-item version of the SCI can be used to rapidly screen for insomnia in routine clinical practice, asking about being troubled by sleep and the frequency of the complaints, with a SCI-02 score of 2 or less indicating insomnia (Table 1). The GP could then assess the insomnia complaint further by administering the remaining six items of the SCI and can also compare the patient’s score to the reference values presented in Table 2 to facilitate clinical interpretation.
Table 2. Two-item version of Sleep Condition Indicator (SCI-2) sex-and age-related reference values, N = 190 000
In conclusion, ultra-short instruments such as the SCI-02 could help GPs and nurses to routinely assess potential sleep problems in their patients, accurately, reliably, and quickly.
Notes
Competing interests
Colin A Espie is co-founder and CMO of Big Health Ltd, which owns the data, and is a shareholder in the company. Annemarie I Luik was employed by the University of Oxford in a post funded by Big Health. Pedro Farias Machado was Head of Data Science with Big Health Ltd, and is salaried by the company. The other author has no conflicts to declare.
- © British Journal of General Practice 2019
