Preserve a missing count when drafting an abstract results table

A draft can turn “the count was not recorded” into “zero participants” when it fills an empty cell with a numeric default. The difference matters before anyone writes a sentence about the observed count.

Here is a synthetic table export. An empty cell means the count is unavailable in this export. A literal zero means the table recorded zero. The script keeps those cases separate.

import csv
from io import StringIO

source = """group,followup_count
A,0
B,
C,8
"""

for row in csv.DictReader(StringIO(source)):
    raw = row["followup_count"].strip()
    label = row["group"]
    if not raw:
        print(f"{label}: count unavailable; check source")
        continue
    count = int(raw)
    print(f"{label}: {count} recorded")

Local output:

A: 0 recorded
B: count unavailable; check source
C: 8 recorded

Group A can keep the literal zero from the export. Group B should stay open until the analyst finds a usable source or explains the missing count. Before writing the final abstract, confirm what the count measures, which population it refers to, and whether the exporter uses a documented token for missing values. Extend the parser for documented tokens only.

This is a constructed example prepared for RigorMed. An AI agent drafted this article for RigorMed and ran the example locally.