Summary of findings for the main comparison. Oral long‐chain omega‐3 PUFAs (ie, EPA and DHA) versus placebo or no treatment.
| Oral long‐chain omega‐3 PUFAs (EPA and DHA) compared with placebo or no treatment for dry eye | ||||||
|
Patient or population: people with dry eyea Settings: primary care setting Intervention: oral long‐chain omega‐3 PUFAs (EPA and DHA)b Comparison: placebo or no treatmentc | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect (95% CI) | No. of participants (studies) | Certainty of the evidence (GRADE) | Comments | |
| Assumed risk | Corresponding risk | |||||
| Placebo or no treatment | Oral long‐chain omega‐3 PUFAs | |||||
| Change in dry eye symptoms, measured using the OSDI score (ranging from 0 to 100 units, with a reduction in scores indicating clinical improvement) at 1‐month (with actual follow‐up ranging from 1 to 12 months) | The pooled summary estimate from four studies (Asbell 2018; Deinema 2017; Goyal 2017; Kangari 2013) indicated little to no reduction in symptoms of dry eye (mean difference [MD] ‐2.47 OSDI units, 95% CI ‐5.14 to 0.19 units, n=677, I2 = 48%). Three additional studies (Bhargava 2015a; Bhargava 2016a; Bhargava 2016b) reported a significant difference in mean symptom score at the end of the follow‐up period, quantified using the Dry Eye Questionnaire and Scoring System (DESS). |
1785 (7 studies) | ⊕⊕⊝⊝ lowd | In addition, symptom score data were reported only as P‐values or in a non‐numeric form in two studies (Kawakita 2013; Pinazo‐Durán 2013), or included non‐dry eye participants (as in Kawashima 2016). | ||
| Change in ocular surface staining at 1‐month (with actual follow‐up ranging from 8 weeks to 12 months) | Two studies found no evidence of a difference between groups for change in corneal fluorescein staining score (Deinema 2017: MD ‐0.31, 95% CI ‐0.66 to 0.04 units, quantified using the Oxford scale; Asbell 2018: MD 0.1, 95% CI ‐0.2 to 0.4 units; P = 0.61, quantified using an unspecified grading scale). Goyal 2017 reported that the control group had a higher rate of conjunctival staining with lissamine green (43.4%) compared with the omega‐3 treatment group (14%), at 3‐months of follow‐up (P=0.009). Kawakita 2013 considered combined corneal and conjunctival staining using rose bengal, and reported that the fish oil (omega‐3) group were “significantly improved to those in the placebo group" after 8 and 12 weeks of treatment. |
681 (4 studies) |
⊕⊕⊝⊝ lowd | No relevant combinable data were available for this outcome. | ||
| Change in Schirmer test (aqueous tear production, measured in mm/5 min, with higher scores indicating more tear production) at 1‐month (with actual follow‐up ranging from 1 to 6 months) | The pooled summary estimate from six studies (Asbell 2018; Bhargava 2015a; Bhargava 2016a; Bhargava 2016b; Deinema 2017; Kangari 2013) showed an improvement in Schirmer test score with long‐chain omega‐3 supplementation relative to the control (MD 0.68, 95% CI 0.26 to 1.09 mm/5 min, n=1704, I² = 16%). In addition, one study (Kawakita 2013, n=26) described no significant inter‐group difference in Schirmer test score, but did not provide quantitative data. |
1730 (7 studies) | ⊕⊕⊕⊝ moderatee | For one additional study, we could not incorporate data into the analyses due to unit of analysis errors (Goyal 2017). Two other trials did not separately report data for dry eye participants versus non‐dry eye (healthy) controls (Kawashima 2016; Pinazo‐Durán 2013). | ||
| Change in tear film stability, measured using tear break‐up time (TBUT) with fluorescein (in seconds, with higher scores indicating greater tear film stability) at 1‐month (with actual follow‐up ranging from 45 days to 12 months) | Four studies (Bhargava 2015a; Bhargava 2016a; Bhargava 2016b; Deinema 2017) reported a significantly improved TBUT with the omega‐3 PUFA intervention compared with placebo, and 1 study (Asbell 2018) reported no significant difference between treatment groups. Meta‐analysis was not performed due to substantial heterogeneity (I² = 98%). | 1640 (5 studies) |
⊕⊕⊝⊝ lowd | The remaining studies were not included in the analysis due to insufficient data reporting (Kangari 2013; Kawakita 2013; Pinazo‐Durán 2013) or the presence of an unit of analysis error (Goyal 2017). | ||
| Change in tear osmolarity (measured in mOsmol/L, with reductions in osmolarity indicating clinical improvement) at 1‐month (with actual follow‐up at 3 months) | Deinema 2017 reported that the mean difference in tear osmolarity was significantly reduced (improved) relative to baseline in the omega‐3 group relative to the placebo group at day 90 (MD: ‐17.71, 95% CI ‐28.07 to ‐7.35 mOsmol/L). | 54 (1 study) |
⊕⊕⊝⊝ lowd | |||
|
Adverse event: gastrointestinal disorders at 1‐month (with actual follow‐up ranging from 3 to 12 months) |
Three studies (Asbell 2018; Bhargava 2016a; Deinema 2017) reported that gastrointestinal disorders were reported in between 5% and 19% of participants in the omega‐3 group and between 0% and 24% of participants in the placebo group. Meta‐analysis was not performed due to substantial heterogeneity (I² = 76%). | 719 (3 studies) | ⊕⊕⊝⊝ lowf | The presence or absence of adverse events was not indicated in three studies (Goyal 2017; Kawakita 2013; Pinazo‐Durán 2013). | ||
| *The basis for the assumed risk (ie, the median control group risk across studies) is provided in the footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; DESS: Dry Eye Questionnaire and Scoring System; DHA: docosahexaenoic acid; EPA: eicosapentaenoic acid; MD: mean difference; OSDI: Ocular Surface Disease Index; PUFA: polyunsaturated fatty acid; RCT: randomized controlled trial; RR: risk ratio; TBUT: tear break‐up time. | ||||||
|
GRADE Working Group grades of evidence.
High quality: further research is very unlikely to change our confidence in the estimate of effect.
Moderate quality: further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.
Low quality: further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.
Very low quality: we are very uncertain about the estimate. aDry eye associated with computer vision syndrome (Bhargava 2015a), rosacea (Bhargava 2016a), visual display terminal use (Bhargava 2016b), and post laser‐assisted in situ keratomileusis (LASIK) (Goyal 2017). The remaining studies included patients with non‐specific causes of dry eye. bDaily dose of EPA and DHA varied substantially between studies, ranging from a daily value of EPA of 85 mg to 2000 mg, and DHA of 108 mg to 1000 mg (see Table 2 for details). cAll studies used a placebo intervention, except for Pinazo‐Durán 2013 (no treatment). dDowngraded one level for each of risk of bias and inconsistency. eDowngraded one level for risk of bias. fDowngraded one level for each of inconsistency and imprecision. | ||||||