Table II.

Distribution of outcome types and intervention comparison types among the 14,886 binary outcome meta-analyses in the Cochrane Database of Systematic Reviews.

	Number (%) of meta-analyses
Outcome types*
Objective outcomes
All-cause mortality	1132 (8%)
Semi-objective outcomes
Obstetric outcomes	1288 (9%)
Cause-specific mortality/major morbidity event/composite (mortality or morbidity)	907 (6%)
Resource use/hospital stay/process	680 (5%)
Surgical/device-related success/failure	623 (4%)
Withdrawals/drop-outs	616 (4%)
Internal/external structure-related outcomes (e.g. radiograph outcomes)	472 (3%)
Subjective outcomes
General physical health indicators (e.g. BMI < 25)	276 (2%)
Adverse events	2330 (16%)
Infection/onset of new acute/chronic disease	2038 (14%)
Signs/symptoms reflecting continuation/end of condition	2184 (15%)
Pain	481 (3%)
Quality of life/functioning (dichotomised)	180 (1%)
Mental health indicators	189 (1%)
Biological markers (dichotomised)†	947 (6%)
Miscellaneous‡	481 (3%)
Intervention comparison types
Pharmacological vs. Placebo/control	5599 (38%)
Pharmacological vs. Pharmacological	4118 (28%)
Non-pharmacological§vs. Placebo/control	2412 (16%)
Non-pharmacological§vs. Pharmacological	315 (2%)
Non-pharmacological§vs. Non-pharmacological§	2442 (16%)

BMI, body mass index.

^**Sixty-two meta-analyses were excluded where the outcome did not fit into any of our pre-defined categories and was classified as ‘Other’.

^††Biological markers (dichotomised) were regarded as subjective outcomes because cutpoints for dichotomisation are often chosen post hoc.

^‡‡Composite (including at least one non-mortality/non-morbidity), satisfaction with care and consumption.

^§§Non-pharmacological interventions: medical devices, surgical, complex, resources and infrastructure, behavioural, psychological, physical, complementary, educational, radiotherapy, vaccines, cellular and gene, and screening.