CPT codes

Sample population

Patients undergoing laparoscopic or open partial colectomy/ileocolectomy with anastomosis (ileocolostomy, colocolostomy, coloproctostomy), using or not using IV fluorescein (or other) injection to evaluate anastomotic vascular supply.

Do you care about the indication?

Outcomes

• Rates of anastomotic leak
• Rates of deep space infection

Do you mean specifically deep infections; deep infections and organ-space infections; or just organ-space infections?

Is this a composite outcome or are these both primary outcomes? Is the composite rate a primary outcome and maybe the two outcomes are secondary endpoint analyses?

Presumably a conditional odds ratio of ICG vs. no ICG in regards to anastomotic leak and rates of deep space infection.

Question

Does the use of intraoperative ICG during colonic anastomosis reconstruction result in lower composite rates of deep space infection and anastomotic leak?

\(H_0\): There is no significant difference in the composite rate of deep space infection and anastomosis leak when ICG is used for intraoperative assessment of a colonic anastomosis.

Build the SQL query

select * from nsqip.puf_cpt

where cpt in ("44160", "44204", "44207", "44145", "44140", "44205") and 
primarysurg is true and
nproc = 1

select * from (
  select *, 
    0 < countif(cpt = '15860' and primarysurg) over(partition by caseid) as icg
  from nsqip.puf_cpt 
)

where cpt in ("44160", "44204", "44207", "44145", "44140", "44205") and 
nproc = 1 and 
primarysurg

select * from (
  select *, 
    0 < countif(cpt = '15860' and primarysurg) over(partition by caseid) as icg
  from nsqip.puf_cpt 
)

join nsqip.puf using(caseid)
join nsqip.col using(caseid, pufyear)

where cpt in ("44160", "44204", "44207", "44145", "44140", "44205") and 
nproc = 1 and 
primarysurg

Cleaning up outcomes

We will simply exclude patients with missing outcome data (rather than imputing it):

dplyr::filter(!is.na(col_anastomotic),
             !is.na(orgspcssi),
             !is.na(wndinfd))