NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.
Observational evidence on the effectiveness of endovascular aneurysm repair compared with open surgical repair of unruptured abdominal aortic aneurysms: Abdominal aortic aneurysm: diagnosis and management: Evidence review K2. London: National Institute for Health and Care Excellence (NICE); 2020 Mar. (NICE Guideline, No. 156.)
Observational evidence on the effectiveness of endovascular aneurysm repair compared with open surgical repair of unruptured abdominal aortic aneurysms: Abdominal aortic aneurysm: diagnosis and management: Evidence review K2.
Show detailsCritical appraisal for this review was performed according to the bespoke instrument shown in Table 4. It amalgamates key criteria from The categories draw on the generic provisions of the Cochrane Collaboration’s ‘Risk Of Bias In Non-randomized Studies of Interventions’ (ROBINS-I) tool (Sterne et al., 2016) and the more technically focused criteria in the ‘Quality of Effectiveness Estimates from Non-Randomised Studies’ (QuEENS) checklist (Faria et al., 2015). See 2.4 for a description of how we developed the instrument.
Table 4Bespoke instrument for appraising risk of bias in casemix-adjusted observational studies
| Criteria | Domain addressed | Risk of bias | Definitions |
|---|---|---|---|
| 1. Selection | |||
| 1.1. Were cohorts from the same time period? | QuEENS 11b | LOW | Contemporaneous recruitment over <5yrs (≥5 yrs if year of operation controlled for) |
| MODERATE | ≥5-yr recruitment with no adjustment for year of operation | ||
| HIGH | Cohorts drawn from different periods in time (e.g. historical controls) | ||
| 1.2. Were cohorts from the same place? | QuEENS 11b | LOW | Same hospital, area or country |
| HIGH | Different sampling frames | ||
| 1.3. Is the definition of AAA the same across cohorts? | QuEENS 11a | LOW | Differences unlikely |
| MODERATE | All infrarenal and complex AAA included, with no adjustment for anatomy | ||
| HIGH | Differences likely (especially where only 1 cohort may include complex AAA) | ||
| 2. Confounding | |||
| 2.1. Does study control appropriately for demographics? |
ROBINS-I 1.4 QuEENS 11c | LOW | At least age and sex |
| MODERATE | Just age or sex | ||
| HIGH | Neither age nor sex, or invalid adjustment methods | ||
| 2.2. Does study control appropriately for comorbidity and/or fitness? |
ROBINS-I 1.4 QuEENS 11c | LOW | Good range of individual comorbidities or validated index (e.g. Charlson, Elixhauser) |
| MODERATE | Limited number of individual comorbidities | ||
| HIGH | None, or invalid adjustment methods | ||
| 2.3. Does study control appropriately for AAA characteristics? |
ROBINS-I 1.4 QuEENS 11c | LOW | At least diameter and some measure of extent (unless study is limited to 1 extent) |
| MODERATE | Just diameter or extent | ||
| HIGH | None, or invalid adjustment methods | ||
| 2.4. Could any adjustment variables have been affected by the intervention? | ROBINS-I 1.6 | LOW | No post-intervention variables controlled for that may mediate treatment effect |
| HIGH | Post-intervention variable(s) controlled for that may mediate treatment effect | ||
| 3. Data collection | |||
| 3.1. Is method of data collection likely to have identified suitable participants accurately? | New | LOW | Medical records |
| MODERATE | Detailed surgical registries (diagnosis and procedure codes specified) | ||
| HIGH | Administrative registries with high-level diagnosis and procedure codes | ||
| 3.2. Is method of data collection likely to record perioperative outcomes accurately? | New | LOW | Medical records or detailed surgical registries |
| HIGH | Administrative registries with high-level diagnosis and procedure codes | ||
| 3.3. Is method of data collection likely to record long-term outcomes accurately? | New | LOW | Direct use of reliable routine data registries |
| MODERATE | Surgical registries or medical records with linkage to reliable routine data registries | ||
| HIGH | Surgical registries or medical records (+/- questionnaires, etc.) | ||
| 4. Analysis – general | |||
| 4.1. Were any checks conducted on model specification and/or fit? | QuEENS 5 | LOW | Residual plots and/or formal tests (e.g. misspecification, autocorrelation, etc.) |
| HIGH | None reported | ||
| 4.2. Are missing outcome data and covariates reported and, if necessary, adjusted for? | ROBINS-I 5.1–5.5 | LOW | Few missing data or amount and types of missingness similar across cohorts |
| HIGH | Possible differential missingness with no valid adjustment or not considered | ||
| 4.3. Have different methods been compared within the study? | QuEENS 1 | LOW | Methods with different assumptions re selection on unobservables compared |
| MODERATE | Different methods compared but all rely on the same assumption about selection | ||
| HIGH | No different methods compared | ||
| 5. Analysis – matching | |||
| 5.1. Is the matching algorithm reported and reasonable? | QuEENS 12–14 | LOW | E.g. nearest-neighbour or caliper/radius matching with reasonable assumptions |
| HIGH | Unreported or invalid methods | ||
| 5.2. Was overlap / common support appropriately assessed? | QuEENS 7 | LOW | Checks on distribution of propensity score with trimming where necessary |
| MODERATE | Comparisons of minima and maxima | ||
| HIGH | No assessment reported | ||
| 5.3. Has balancing of the covariates been demonstrated? | QuEENS 8 | LOW | Normalised differences in covariates reported, with none >0.25 |
| MODERATE | Conventional hypothesis tests, with no evidence of significant differences | ||
| HIGH | Meaningful differences in 1 or more important covariates or not reported | ||
| 6. Analysis – simple multivariable models | |||
| 6.1. Is sample size adequate relative to number of covariates considered? | New | LOW | Number of events is ≥10 times greater than number of variables considered |
| HIGH | Number of events is <10 times greater than number of variables considered | ||
| 6.2. Were interactions between treatment and other covariates considered? | New | LOW | Yes |
| HIGH | No | ||
| Overall | |||
|
Domain scores represent the average score given for each question (rounded mean where 1=low, 2=medium & 3=high) Overall judgement based on rules stated here. | LOW | 0 domains high and <2 moderate | |
| MODERATE |
0 domains high and >1 moderate or 1 domain high and <2 moderate | ||
| HIGH |
>1 domain high or 1 domain high and >1 moderate | ||
- Instrument for appraising risk of bias - Observational evidence on the effective...Instrument for appraising risk of bias - Observational evidence on the effectiveness of endovascular aneurysm repair compared with open surgical repair of unruptured abdominal aortic aneurysms
- Evidence tables for included studies - Observational evidence on the effectivene...Evidence tables for included studies - Observational evidence on the effectiveness of endovascular aneurysm repair compared with open surgical repair of unruptured abdominal aortic aneurysms
- RecName: Full=Lactoylglutathione lyase; AltName: Full=Aldoketomutase; AltName: F...RecName: Full=Lactoylglutathione lyase; AltName: Full=Aldoketomutase; AltName: Full=Glyoxalase I; Short=Glx I; AltName: Full=Ketone-aldehyde mutase; AltName: Full=Methylglyoxalase; AltName: Full=S-D-lactoylglutathione methylglyoxal lyasegi|134039205|sp|Q04760.4|LGUL_HUMANProtein
- Taxonomy Links for Nucleotide (Select 87239991) (1)Taxonomy
- PubChem Compound Links for Gene (Select 56386) (28)PubChem Compound
Your browsing activity is empty.
Activity recording is turned off.
See more...