Table 1. Content validation indices–DRAFT ResQu items, proportion of expert panel giving a positive rating*.
| Item | Clarity | Relev-ance | Import-ance | |
|---|---|---|---|---|
| 1 | Defines and describes each birth settings clearly | .90 | 1.00 | 1.00 |
| 2 | Type of study design | .90 | .90 | .95 |
| 3 | Uses reliable and logical comparison group/s | .76 | .95 | 1.00 |
| 4 | Retains women in original birth setting cohort for data analysis (intention to treat) | .95 | 1.00 | 1.00 |
| 5 |
Distinguishes between
a) planned home births with skilled attendants AND b) free births or unplanned home births (if home births included in study) |
.95 | 1.00 | 1.00 |
| 6 | Identifies planned place of birth at time in pregnancy that is appropriate to selected outcome measures | .90 | 1.00 | .95 |
| 7 | Accounts for effect of provider type | .67 | .90 | .81 |
| 8 | Discriminates outcomes of care according to type of provider (as distinct from birth setting) | .80 | .81 | .71 |
| 9 | Sample size powered appropriately for selected outcomes being measured | .90 | .90 | .90 |
| 10 | Uses reliable method of initial sampling and recruitment for each cohort | .71 | .95 | .90 |
| 11 | Provides consistent inclusion criteria for comparison groups | .86 | .95 | .90 |
| 12 | Uses reliable method to track women when birth setting changes | .90 | 1.00 | 1.00 |
| 13 | Addresses effect of level of service integration between home, birth centre and hospital | .86 | .86 | .90 |
| 14 | Controls for confounders including socio-demographic and health profile of women in cohorts | 1.00 | 1.00 | 1.00 |
| 15 | Reports criteria for transfer (change of birth place) | 1.00 | .90 | .62 |
| 16 | Considers potential effects related to timing of transfer and delays to treatment | .81 | .95 | .86 |
| 17 | Accounts for effect of mode of transfer (ambulance, private car, neonatal transport team etc.) | .71 | .62 | .48 |
| 18 | Defines key terms (e.g. PPH, low risk, planned home birth, mortality, morbidity) consistently and transparently using recognised methods and definitions (e.g. NICE, RANZCOG or ACOG guidelines) | .95 | .95 | .95 |
| 19 | Applies reliable statistical methods to compare cohorts, e.g. absolute risk, relative risk, confidence intervals | .85 | .95 | .95 |
| 20 | Reports and minimises missing data | .95 | .95 | .90 |
| 21 | Draws conclusions based on reported data | 1.00 | .90 | .90 |
| 22 | Acknowledges impact of lack of randomisation | .95 | .76 | .76 |
| 23 | Acknowledges impact of size of cohorts for each outcome measured | 1.00 | .90 | .76 |
| 24 | Acknowledges impact of retrospective and/or incomplete data | .90 | .90 | .81 |
| 25 | Acknowledges impact of local/regional standards, policies and protocols | .95 | .95 | .71 |
| S-CVI: Average (mean) of I-CVIs of retained items | .89 | .94 | .90 | |
| S-CVI: Proportion of retained items with expert-rated I-CVIs of >80% | .82 | 1.00 | .90 | |
| REVISIONS: | ||||
| Items 15, 17 and 22 (shaded) deleted. | ||||
| Items 7 and 8 collapsed into one item (item 10 Fig 1) | ||||
| Item 19 split into two (items 15 and 21 in Fig 1) | ||||
| Several items reworded in line with comments from experts in survey or correspondence | ||||
* Positive rating is either 1 “very clear/important/relevant” or 2 “clear/important/relevant but needs minor revision”