/data · research mirror

Lab dashboards. Anaemia. Tuberculosis.

A cross-sectional view of the diagnostic cohort at Anubhav Life Care, Barasat, with anaemia and tuberculosis treated as separate research domains. Each chart relates a result back to the rest of the panel or to patient demographics — gender is inferred from given name where it was not captured at registration.

Illustrative figures · scaled to the AKTIV → Neon research mirror · cohort n ≈ 18,400 patients · 36 curated TEST_KEYs · cross-sectional only

01 · anaemia

Anaemia × the rest of the panel

Cohort: patients with at least one haematology or iron-studies result in the curated anaemia TEST_KEY list. Severity follows WHO Hb cutoffs adjusted for sex.

cohort n = 9,842rolling 36 months

Cohort demographics

gender from name where unrecorded

9,842

patients

57 : 43

F : M

AGE DISTRIBUTION (years)

5–14

15–24

25–44

45–64

65+

Anaemia severity

WHO Hb thresholds, by sex

Normal5,47155.6%
Mild2,63026.7%
Moderate1,35213.7%
Severe3894.0%

Anaemia prevalence by sex × age

% with Hb below WHO cutoff

<55–1415–2425–4445–6465+

Female

62%

51%

58%

54%

41%

47%

Male

38%

33%

22%

19%

26%

38%

Hb vs MCV — morphology

microcytic ←→ macrocytic, n = 4,371 anaemic

Iron studies (anaemic patients)

serum Iron / TIBC / ferritin medians

Serum Iron (μg/dL)42 (ref 60–170)
TIBC (μg/dL)418 (ref 250–450)
Transferrin sat (%)11 (ref 20–50)
Ferritin (ng/mL) F14 (ref >15)
Ferritin (ng/mL) M28 (ref >30)

B12 deficiency overlap

anaemic patients with paired B12 result, n = 1,184

Microcytic
12%
85%
Normocytic
18%
73%
Macrocytic
47%
24%
29%

B12 < 200 pg/mL
200–300
≥ 300

Anaemia × HbA1c (diabetes)

anaemic patients with paired HbA1c, n = 1,612

HbA1c (%)

<5.75.7–6.46.5–7.9≥8.0

Normal

58%

24%

12%

Mild

49%

27%

16%

Moderate

41%

28%

19%

12%

Severe

33%

26%

22%

19%

row: Anaemia

Hb vs Ferritin (log)

n = 2,408 paired results, Spearman ρ = 0.47

02 · tuberculosis

Tuberculosis screening results

Cohort: patients with at least one sputum AFB, AFB-PCR, mycobacterial culture, Mantoux, or IGRA result in the curated TB TEST_KEY list. Active vs latent is not always separable — both views are kept here.

cohort n = 3,186rolling 36 months

Cohort demographics

referred for any TB workup

3,186

patients

44 : 56

F : M

AGE DISTRIBUTION (years)

5–14

15–24

25–44

45–64

65+

Sputum AFB result

ZN smear · n = 1,840

Negative1,58286.0%
Scanty985.3%
1+844.6%
2+472.6%
3+291.6%

Modality concordance

patients with paired results

SmearAFB-PCRCultureIGRA

Smear

258

162

138

AFB-PCR

162

174

151

Culture

138

151

153

IGRA

214

cells = patients positive on BOTH modalities · diagonal = single-modality positives

Mantoux induration (mm)

n = 612, IGRA-paired subset shaded

218

0–4

134

5–9

121

10–14

15–19

20+

allIGRA+induration (mm)

TB-Gold (IGRA) result

QFT-Plus · n = 738

Negative50268.0%
Positive21429.0%
Indeterminate223.0%

Positivity by age × sex

smear, PCR, culture or IGRA positive

<55–1415–2425–4445–6465+

Female

11%

14%

12%

Male

16%

22%

24%

18%

TB × anaemia comorbidity

TB-positive patients with paired Hb, n = 412

Smear +ve
21%
34%
31%
14%
PCR/culture +ve
24%
36%
28%
12%
IGRA +ve only
49%
28%
17%

Hb normal
Mild anaemia
Moderate
Severe

TB × diabetes comorbidity

TB-positive patients with paired HbA1c, n = 318

HbA1c (%)

<5.75.7–6.46.5–7.9≥8.0

Smear +ve

44%

22%

18%

16%

PCR/culture +ve

41%

24%

19%

16%

IGRA +ve only

62%

21%

11%

02b · TB pathology & vitals

Pathology and vitals signature in TB

From the patient-level cohort: every CBC, biochemistry, urine, body-fluid and (where paired) SamaClip vitals draw for any patient who ever had a TB workup, regardless of which day the sample was collected. TB-positive = smear / PCR / culture / IGRA reactive on any visit.

CBC signature — TB+ vs TB−

median of patient's most recent CBC

TB +veTB −vemedian value

Hb (g/dL)ref 13.0 / 12.0
10.4
12.6
WBC (×10⁹/L)ref 4–11
9.8
7.6
Neutrophils (%)ref 40–75
71
58
Lymphocytes (%)ref 20–45
18
32
Monocytes (%)ref 2–8
10.4
5.2
Monocyte : Lymph ratioref < 0.32
0.58
0.16
Platelets (×10⁹/L)ref 150–410
412
268
ESR (mm/hr)ref < 20
78
18

Elevated monocyte-to-lymphocyte ratio and reactive thrombocytosis remain the most consistent CBC fingerprint of active TB in this cohort.

Serum chemistry signature — TB+ vs TB−

median of latest paired draw

TB +veTB −vemedian value

CRP (mg/L)ref < 5
48
5
Albumin (g/dL)ref 3.5–5.0
3.1
4.2
Total protein (g/dL)ref 6.4–8.3
7.8
7.2
A : G ratioref 1.1–2.0
0.82
1.41
ALT (U/L)ref < 40
31
22
AST (U/L)ref < 40
38
24
Sodium (mmol/L)ref 135–145
132
138
Vitamin D (ng/mL)ref ≥ 30
14
22
Ferritin (ng/mL)ref 30–300
248
96
Serum ADA (U/L)ref < 24
38
14

Low albumin + hyponatraemia + high CRP + low vitamin D is the classic chronic-infection picture; raised serum ADA is the cohort-specific finding worth tracking.

Urinary findings — TB+ patients

latest urine, n = 482

Pyuria (≥10 WBC / HPF)34% (164/482)
Microscopic haematuria21% (101/482)
Proteinuria ≥ 1+28% (135/482)
Leukocyte-esterase +ve31% (149/482)
Nitrite +ve9% (43/482)
Urine AFB +ve (when sent)4% (4/89)
Sterile pyuria (pyuria w/o organism)14% (67/482)

Sterile pyuria with a negative routine culture is the urinary tract's classic flag for genitourinary TB. Sent urine for AFB on 89 / 482 — yield was 4 %.

Body-fluid analysis — TB-suspected taps

paired biochem + cytology, where sent

Pleural fluidn = 94

Protein > 3.0 g/dL (exudate)
91%
ADA > 40 U/L
67%
Lymphocyte predominance (> 50 %)
78%
LDH ratio (fluid : serum) > 0.6
72%
AFB smear +ve
8%

CSFn = 32

Protein > 0.45 g/L
84%
Glucose ratio < 0.5
69%
ADA > 10 U/L
53%
Lymphocytic pleocytosis
72%
AFB / PCR +ve
25%

Ascitic fluidn = 41

SAAG < 1.1 (exudate)
71%
ADA > 39 U/L
61%
Lymphocyte predominance (> 50 %)
66%
Protein > 2.5 g/dL
78%
AFB / culture +ve
17%

Non-invasive vitals overlay

SamaClip + lab paired subset, n = 218

TB +veTB −vemedian value

SpO₂ at rest (%)ref ≥ 95
95.2
98.1
Heart rate (bpm)ref 60–100
96
78
Respiratory rate (PPG proxy)ref 12–20
22
16
HRV — RMSSD (ms)ref ≥ 25
19
34
Non-invasive Hb (g/dL)ref 13.0 / 12.0
10.7
12.8
QTc on screening ECG (ms)ref < 450
432
408

Where a SamaClip screen was paired to the lab visit (n = 218), TB+ patients sit lower on SpO₂, breathe faster, have suppressed HRV, and trend toward QTc-prolongation worth monitoring on bedaquiline / moxifloxacin therapy.

Pathology hits per TB+ patient

ever-positive cohort, n = 412

paired-positive finding

Anaemia↑ CRPHyponat.↓ Vit DPyuria↑ Mono:Lymph

Smear +ve

79%

88%

41%

72%

38%

81%

PCR/culture +ve

74%

84%

38%

68%

36%

76%

IGRA +ve only

52%

49%

18%

58%

22%

44%

row: modality

methods

How the numbers were assembled

Source. AKTIV laboratory information system at Anubhav Life Care, mirrored nightly into a research-only Neon Postgres project. The cohort is defined by a curated set of 74 anaemia / tuberculosis / cardiac TEST_KEYs, but once a patient is in the cohort the mirror pulls the full pathology panel for every visit they ever had — haematology, biochemistry, hormones, serology, microbiology, urine, body fluids, cytology, sputum / culture / Mantoux / IGRA, and radiology free-text reports — so that comorbidities, complications, and incidental findings can be related back to the index condition.

Demographics. Age from MAST_PATIENT.DOB; sex from MAST_PATIENT.SEX where present, otherwise inferred from given name against a Bengali / Hindi / English first-name dictionary. Records where neither field nor name resolves are reported as unknown and excluded from sex-stratified panels.

Definitions. Anaemia cutoffs follow WHO (Hb < 13.0 g/dL adult M, < 12.0 non-pregnant adult F, < 11.0 pregnant F, < 11.5 children 5–11). TB-positive means any of: sputum AFB ≥ scanty, AFB-PCR detected, culture growth of M. tuberculosis, or IGRA reactive.

Vitals overlay. Where a patient also had a SamaClip non-invasive screen on the same day as a lab draw, the screen's outputs — non-invasive Hb, SpO₂, heart rate, PPG-derived respiratory rate, HRV (RMSSD), and screening ECG QTc — are joined to that lab draw on PATIENT_KEY × draw date. Only paired rows are counted in the vitals tile.

Caveat. Cross-sectional only. A single patient with repeat draws contributes once per panel, using the most recent value. Longitudinal trends are out of scope for this page.