Extraction form for project: SGS 2022 Obesity associations

Design Details

1. Include or Reject
2. Rejection reason
Confirm none of these is a reason for rejection. If you select a rejection reason, stop and move on to next article
3. Obesity threshold
In the multivariable analysis, what was the threshold for obesity? If the analysis included multiple thresholds, you can check multiple options. If "Other", enter the threshold (a single number) in the follow-up question.
4. Mesh
Did the study involve use of mesh?
5. Study design
The underlying study design
6. Do the data come from a registry or other large/major database? If so, which?
7. Directionality
Prospective = data collected based on a study protocol. (Nested case-control are generally prospective. Registries are generally retrospective, even if data entered into registry "prospectively".) If unclear, explain why.
8. Timepoints
Longitudinal = outcomes followed after surgery/hospitalization. Cross-sectional = one time point (which may be the perioperative period).
9. Population/Condition
If >10% of women had condition other than prolapse or incontinence, confer with Nancy before continuing.
10. Prolapse surgery category
11. Incontinence surgery category
12. Study dates
Calendar years of surgery/procedure/inclusion in study. Just the years. If unclear, you can enter "<" the publication year (eg, <2105 for study published in 2015); in this case, leave "Earliest" blank.
Earliest
Latest
13. Study location
Country or region or city. Select one or more options and enter data. Especially where multicenter, no need to include every detail (eg, every city). Can include a summary description (e.g., "Across the East Coast"). Use your judgment what will be sufficiently informative.
14. Setting
Excuse any redundancy with Q12.
15. Centers
16. Hospital type
17. Eligibility: BMI
Describe restrictions on eligibility based on BMI. If they EXCLUDED "underweight" (eg, BMI <18), in "Lowest weight included" select 18; leave "Highest weight included" blank. If they INCLUDED BMI 25-40, in "Lowest weight included" row select 25; in "Highest weight included" row select 40. Don't worry about < vs. <= (or > vs. >=).
No exclusions based on BMI (explicit)
No exclusions based on BMI (implicit, NR)
Lowest BMI included
Highest BMI included
18. Eligibility criteria
Other than BMI [You captured this in the prior question]. Be concise; use common abbreviations. Omit nonclinical/demographic factors (like consent). Do not be redundant between inclusion and exclusion. (It's okay to convert their exclusions into inclusions; for example, instead of "we excluded children <18, it will make for better consistency across studies to extract as an inclusion that they included adults >=18). It's okay to leave exclusions blank.) If you copy and paste, please clean up the resulting typos and line breaks.
Inclusion
Exclusion
19. Lack of generalizability (atypical population)?
Is this a "special" or "focused" or "atypical" population? If yes, succinctly describe (eg, inner-city Blacks, ...). Use your judgment as to what would be an interesting "special" or highly restricted population that Nancy should be aware of.
20. Sample size
Total eligible = sample size after eligibility exclusions (not including dropout due to missing data). Model N = number included in final model. Enter NR if either is not reported.
Total eligible
Model N
21. BMI
Note that full range might be derived from eligibility criteria. Enter data for the whole sample only. For ranges, put commas between numbers. FOR THIS AND ALL BASELINE CHARACTERISTICS, if the article reports only by intervention (or by group), please calculated the weighted average ((N1*mean1)+(N2*mean2))/(N1+N2). Enter NR for SD.)
Full range
Mean (SD)
Median (IQR)
Not reported
22. Age
Note that full range might be derived from eligibility criteria. Enter data for the whole sample only. For ranges, put commas between numbers.
Full range
Mean (SD)
Median (IQR)
Not reported
23. Parity
Note that full range might be derived from eligibility criteria. Enter data for the whole sample only. For ranges, put commas between numbers.
Full range
Mean (SD)
Median (IQR)
Not reported
24. Race/Ethnicity
Enter % (0%-100%); NOT proportion (0-1). For "Asian (specific)" or "Other" enter race/nationality and %. If there are multiple, enter all together, with commas between. If study is from Asia, you may be able to assume they are 100% Asian (eg, Japanese). Leave rows with NR blank, but if overall NR then select that option at the bottom of the table.
White (any)
White, non-Hispanic
Black (any)
Black, non-Hispanic
Non-White
Non-Black
Hispanic (any)
Hispanic, White
Hispanic, Black
Asian (any)
Asian (specific)
Native/Indigenous
Other
NR
25. Comorbidities etc.
Enter % (0-100; not proportion) with each comorbidity. If NR, leave blank (except if overall NR, then check this in last row). Enter Other (comorbidity and %) only if a particularly pertinent factor was omitted from our list. Let Nancy know (now) for possible explicit inclusion in the extraction form.
Diabetes
CHF
HTN
Smoking
Post-menopausal
Diuretic use
Other
NR
26. Regression method
If linear regression, confer with Nancy and Ethan before continuing. If Other/unclear, explain and confer with Nancy and Ethan before continuing.
27. Outcome analyzed
If outcome category is Other, check with Nancy, before proceeding. If specific outcome is Composite complications, specify whether UTI is included.
Outcome category
Specific outcome
Follow-up duration
Comment/Note
28. Was complete list of association estimates reported for all analyzed covariates (both significant and nonsignificant)?
In other words, is there an OR or aOR (or RR or HR or beta) for all analyzed covariates, not only the factors in the final model? FOR THIS, AND ALL RELEVANT SUBSEQUENT QUESTIONS, EXTRACT THE FULLEST MODEL (unless they include inappropriate covariates and a less-full model excludes these). Feel free to check with Ethan if you can't decide. Add comment only if necessary.
Y/N
Comment/Note
29. Was final model reported in full?
This question does not refer to variables omitted from the model. If they say something like "adjusted for x, y, z" without reporting numbers for x, y, z, then No. If they say only that the adjusted OR for obesity is X (without listing all the other variables in the model and their estimates), then No. If the answer to Q25 is Yes, then the answer to this question is Yes. Add comment only if necessary.
Comment
30. Inappropriate covariates
Did model include covariates measured after/during surgery (like EBL, LOS, complications, post-op incontinence)? Or any other questionable covariates (feel free to check with Ethan)? If yes, list these (AND STOP HERE).
31. Number of covariates
How many covariates are included in the final model? Include interaction terms and listed covariates that are not reported explicitly). Do not include the intercept. If univariate, enter 1.
Number
Cannot determine
32. Estimate 1
Each comparison of BMI categories entered separately. If they divided into 3 groups (eg, <25, 25-30, >30) and used <25 as the reference group, they should have two rows of estimates/data. One for 25-30 vs. <25 and one for >30 vs. <25. The 25-30 vs. <25 will go here as "Estimate 1". The >30 vs. <25 will go in the next question (Estimate 2), etc. Enter the reference group (<25) for each Estimate. Succinctly describe the BMI categories being analyzed (eg, <25, 25-30). Under n/N, check radio button then enter data in text box that appears. Calculate the % (not the proportion) as needed; the denominator is the number in the model. For n/N and %, enter "NR" if the case. Please enter the estimate (eg, OR) in the form E.EE (L.LL, U.UU) with spaces between items and a comma not a dash. REMEMBER, we are including univariate estimates when BMI categories were tested and excluded from final model for being NS. If they report only NS (without an estimate), enter NS under P value and check NR in the second column of the Estimate row.
Definitionn/N%
BMI category (evaluated)
BMI category (reference)
Measure
Estimate (95% CI): E.EE (L.LL, U.UU)
P value
Univariate?
33. Estimate 2
Each comparison of BMI categories entered separately. If they divided into 3 groups (eg, <25, 25-30, >30) and used <25 as the reference group, they should have two rows of estimates/data. One for 25-30 vs. <25 and one for >30 vs. <25. The 25-30 vs. <25 will go here as "Estimate 1". The >30 vs. <25 will go in the next question (Estimate 2), etc. Enter the reference group (<25) for each Estimate. Succinctly describe the BMI categories being analyzed (eg, <25, 25-30). Under n/N, check radio button then enter data in text box that appears. Calculate the % (not the proportion) as needed; the denominator is the number in the model. For n/N and %, enter "NR" if the case. Please enter the estimate (eg, OR) in the form E.EE (L.LL, U.UU) with spaces between items and a comma not a dash. REMEMBER, we are including univariate estimates when BMI categories were tested and excluded from final model for being NS. If they report only NS (without an estimate), enter NS under P value and check NR in the second column of the Estimate row.
Definitionn/N%
BMI category (evaluated)
BMI category (reference)
Measure
Estimate (95% CI): E.EE (L.LL, U.UU)
P value
Univariate?
34. Estimate 3
Each comparison of BMI categories entered separately. If they divided into 3 groups (eg, <25, 25-30, >30) and used <25 as the reference group, they should have two rows of estimates/data. One for 25-30 vs. <25 and one for >30 vs. <25. The 25-30 vs. <25 will go here as "Estimate 1". The >30 vs. <25 will go in the next question (Estimate 2), etc. Enter the reference group (<25) for each Estimate. Succinctly describe the BMI categories being analyzed (eg, <25, 25-30). Under n/N, check radio button then enter data in text box that appears. Calculate the % (not the proportion) as needed; the denominator is the number in the model. For n/N and %, enter "NR" if the case. Please enter the estimate (eg, OR) in the form E.EE (L.LL, U.UU) with spaces between items and a comma not a dash. REMEMBER, we are including univariate estimates when BMI categories were tested and excluded from final model for being NS. If they report only NS (without an estimate), enter NS under P value and check NR in the second column of the Estimate row.
Definitionn/N%
BMI category (evaluated)
BMI category (reference)
Measure
Estimate (95% CI): E.EE (L.LL, U.UU)
P value
Univariate?
35. List of covariates in model
List the covariates in the model (other than obesity). Please simplify down to single words as possible (eg, "age", not a list of each of the 3 age categories they included). PUT A SEMICOLON (;) BETWEEN EACH COVARIATE.
36. 2nd Review
For 2nd reviewer: List corrections (from your perspective) pertinent to the extracted data. You don't need to list trivial corrections (like typos; except maybe number typos).
Reviewer name
Non-trivial corrections made
Confirmed with extractor and/or Nancy
37. Was the comparison of interventions/procedures included as a variable in the regression model?
If, for example, they compared 2 surgical approaches, the regression model should have a covariate for surgical approach (whether significant or not).
38. Comments/Notes

Risk of Bias Assessment

1. Were eligibility criteria clear?
Comment if No
Rating
Notes/Comments:
2. Were interventions adequately described?
Comment if No
Rating
Notes/Comments:
3. Were the outcomes fully defined?
Comment if No
Rating
Notes/Comments:
4. Clear reporting with no discrepancies
Comment if No
Rating
Notes/Comments:
5. Statistical methods adequately conducted and described
Comment if No
Rating
Notes/Comments:
6. Analyzed rate >80 percent
Is No. included in model divided by No. in database> 80%? What % if No
Rating
Notes/Comments:
7. No. analyzed ÷ No. variables >10
Rating
Ratio
8. Full accounting of all analyzed variables
Did the article list/describe the full list of variables that were assessed for possible inclusion in the model. Or did they fully list all variables in the model AND all nonsignificant variables that were excluded from the model? Comment if No
Rating
Notes/Comments:
9. 2nd Review
Changes made
Confirmed with extractor and/or Nancy