Discrepant Sahli haemoglobin results before an antenatal camp
The only laboratory at a primary health centre in a block of eastern Uttar Pradesh is staffed by a single multipurpose health worker (lab). On Tuesday morning, 40 antenatal cards have to be completed before a maternal health camp begins at 11 a.m. The Sahli haemoglobinometer is the only method available. The N/10 hydrochloric acid was prepared three weeks ago and is stored in a glass bottle; the comparator tubes are noticeably faded; and a small clot is visible near the tip of the 20 µL pipette. Three results from the same morning: a 24-year-old primigravida reads 6.8 g/dL; a 30-year-old in her third trimester reads 7.2 g/dL; a 26-year-old whose card shows 11.6 g/dL six weeks ago now reads 7.8 g/dL, although she looks well and has no pallor, breathlessness or tachycardia. The medical officer is in the labour room and wants the reports by noon. There is no automated analyser, no commercial control blood, and no time to send samples to the district laboratory before the camp.
1. Which of the three haemoglobin values should you treat as least reliable, and how would you work out why before releasing any report?
Read the worked answer
Least reliable: the 7.8 g/dL in the 26-year-old. A genuine fall of 3.8 g/dL in six weeks would almost certainly produce visible pallor, fatigue, tachycardia or breathlessness; she has none of these. The result also conflicts with her previous value from the same method, so it is a discordance, not a trend.
Reasoning and stepwise check:
- Pipette: a clot near the tip means less than 20 µL of blood is drawn, which dilutes the acid-haematin colour and gives a falsely low reading. This is the most likely single cause.
- Reagent age and storage: N/10 HCl more than about a week old, or stored in a poorly stoppered bottle, can concentrate by evaporation or become contaminated, altering conversion and colour.
- Comparator: faded standards make visual matching unreliable in both directions, and the error grows at the anaemic end of the scale.
- Technique and timing: reading before full colour development (about 5–10 minutes) or in poor light biases results.
What to do: do not release the 7.8 g/dL as final. Clean or replace the pipette, prepare fresh N/10 HCl, check the comparator against distilled water, and repeat the sample in duplicate in good daylight. If the repeat agrees with the clinical picture and with a second sample, release it with a note; if it stays low, refer her for confirmation at the district laboratory rather than entering a value you cannot defend.
2. What internal quality control can you run at the same sitting without commercial control material, and what acceptance criteria would you apply?
Read the worked answer
Sahli's method has no true calibrator, so control must come from consistency checks and cross-comparison.
- Duplicate testing: run the same blood twice from a fresh finger prick, ideally by two people or with two pipettes. Sahli visual matching is imprecise; accept a difference of ≤0.5 g/dL between duplicates. Wider spread means the method or the operator is out of control today.
- Reagent blank check: compare the unused acid tube and a distilled-water-filled tube against the comparator to detect a faded or mismatched set; a visible mismatch means the set should be replaced.
- Biological reference: test one healthy, non-anaemic staff volunteer whose Hb is expected to be 12–15 g/dL. A value far outside that range signals a systematic bias, not a patient problem.
- External comparison: send two samples (one low, one normal) in EDTA to the district laboratory and compare. A consistent offset of 0.5–1 g/dL is known for visual Sahli against cyanmethaemoglobin, but a variable offset is not acceptable.
- Record everything: date, reagent batch, comparator set, duplicate results, action taken. A control that is not written down has not been done.
If the duplicate criterion fails after correction, stop reporting haemoglobin for the day, inform the medical officer, and arrange referral testing until the set is repaired or replaced.
3. What should be recorded and communicated, and how should the 26-year-old's antenatal card be handled?
Read the worked answer
- Register: enter all three results with a clear remark for the discrepant one - 'repeat required, previous 11.6 g/dL, clinically well' - so that anyone reading later sees the doubt rather than a clean false number.
- Antenatal card: do not write 7.8 g/dL as a confirmed value. Write the repeat result once obtained, or note 'Hb repeat advised, report awaited'. A card that carries a wrong low value may trigger unnecessary iron loading, anxiety or referral.
- Inform the medical officer: a short verbal or phone message with the reason, so that clinical decisions at the camp are not based on unverified data. If a genuine reading is below 7 g/dL, that woman needs prompt evaluation and likely referral regardless of the camp schedule.
- Patient communication: explain simply that the machine reading did not match her earlier report and that a repeat is needed; ask her to wait or return the same day. Do not alarm her.
- Quality records: note the corrective action in the laboratory QC log and flag the pipette, acid and comparator for replacement. If this is the second such event, report it upward as an equipment/reagent problem rather than repeatedly blaming technique.
- Recall: keep a short list of patients needing a repeat so that none are lost after the camp.