twelve times

too many patients

Problem statement

India's National Sickle Cell Anaemia Elimination Mission (launched July 2023) set out to screen 70 million people aged 0–40 in tribal and high-prevalence districts by 2025–26, issue each a status card, counsel carriers and treat patients; by July 2025 it had screened 60 million and reported 215,000 people with sickle cell disease and 1.67 million carriers, with 26 million cards distributed. Screening in the field relies on the solubility sickling test or one of the rapid point-of-care tests (POCTs) that ICMR has validated — but the solubility test cannot distinguish a person with the disease (SS) from a healthy carrier (AS), two of the POCTs on ICMR's approved list were validated on just 35 and 70 samples with claimed 100% sensitivity and 100% specificity (sample size "to be increased") and most kits approved in 2023–24 report 100%/100% on 120–300 samples, no current POCT can identify sickle–β-thalassaemia, and confirmatory HPLC is, in the ICMR-NIIH co-authored commentary's words, "possibl[y] … not systematically conducted." The result, noted by an ICMR-NIIH scientist and colleagues, is that the number of disease cases recorded is "almost 12-fold larger" than Hardy–Weinberg genetics would predict from the carrier count. The commentary itself points first to consanguinity (an inbreeding coefficient of ~0.021 from the dashboard) and "a complex range of factors"; whether the gap is misclassification of carriers as patients, population structure and consanguinity, or ascertainment, nobody can currently say — and 60 million people have been given cards on the strength of it.

Why this matters

A card that says "disease" when the person is a carrier — or "carrier" when the person has disease — has lifelong consequences in the very communities the mission targets: it drives marriage counselling, prenatal decisions, hydroxyurea and follow-up eligibility, and stigma, and India's Scheduled Tribe population (67.8 million by the 2011 census) is already, in the mission portal's words, disproportionately burdened. At programme scale, misclassification distorts the epidemiology on which district resourcing, drug procurement and the 2047 elimination target rest, and it undermines confidence in a screening infrastructure that could otherwise be a global model. The mission's own portal frames the ambition — universal population screening, "a strong network of diagnosis and linkages" and "robust monitoring" — so the gap between screen-positive and confirmed diagnosis is a stated aim left unmet, not an afterthought.

What’s been tried and why it hasn’t worked

ICMR built a kit-validation pathway (ICMR-NIIH Mumbai, ICMR-CRMCH Chandrapur, ICMR-RMRC Bhubaneswar and others) and approved a growing list of POCTs, which is what made mass screening feasible; but the validation sample sizes range from 1,559 (HemoTypeSC, 2018–20) down to 35 and 70 for kits approved in 2023 with perfect claimed performance, and the ICMR list itself notes that "appropriate training of the field workers is required before use of the PoCT for avoiding errors in interpretation." The gold-standard confirmation, HPLC, "tends to be expensive and relies on highly skilled staff, sophisticated equipment requiring regular maintenance, and available reagents," so screen-positive results are frequently not confirmed and the solubility test — which reads any HbS as positive — remains in use because it is cheap. Newborn screening by solubility also suffers high false-negative rates because of fetal haemoglobin. Field POCTs that do separate AS from SS exist (lateral-flow immunoassays), but no independent, protocol-standardised evaluation across kits and across India's tribal populations (where HbS coexists with β-thalassaemia and other variants) has been done, and no public data yet report what fraction of card-holders were confirmed.

What would unlock progress

Two things would resolve the ambiguity: an inexpensive, scalable confirmatory tier between POCT and HPLC (batched capillary electrophoresis, dried-blood-spot HPLC hubs, or a second-line POCT with independently validated AS/SS discrimination), and a data-quality layer that uses the mission's own dashboard to flag districts, kits or teams whose disease:carrier ratios are genetically implausible and route those records to confirmation. Adjacent precedent: HIV programmes moved from single rapid tests to serial-testing algorithms with defined confirmatory rules and external quality assessment; newborn-screening programmes in the US and Brazil pair a cheap first-tier test with mandatory second-tier confirmation and track positive predictive value per site.

Entry points for student teams

A data team could take the published state- and district-level mission counts and build a Hardy–Weinberg / population-structure plausibility model that estimates the expected disease-to-carrier ratio under stated assumptions (allele frequency, consanguinity, targeted-district selection) and quantifies where the recorded ratio is inconsistent — a proof-of-concept audit tool the mission could run monthly. A diagnostics team could write the head-to-head evaluation protocol the ICMR list is missing — comparator, per-kit sample sizes powered to separate 97% from 100% sensitivity rather than to confirm it, mandatory inclusion of HbS/β-thalassaemia compound heterozygotes and HbF-heavy newborn specimens, blinded reading, and per-site positive predictive value reporting — anchored on what the two recent multicentre evaluations already measured (Shrestha et al. compared six low-cost tests against HPLC in 138 participants across Nepal and Canada, Lancet Reg Health Southeast Asia 2025; Thaker et al. put HemoTypeSC against HPLC in 1,725 Indian newborns across six centres and got 93.3% sensitivity, not the 100% the kit list claims, Indian J Med Res 2025) and handed to ICMR-NIIH or a mission district lab to run, since newborn and tribal-district specimens require Indian ethics approval and a partner institution that owns the samples. A health-systems team could design the confirmatory pathway (sample logistics, hub-and-spoke HPLC, card-issuance rules) and cost it per screened person. Relevant skills: population genetics and statistics, laboratory diagnostics, health-systems design and implementation science.