Diagnostic Lab — Malaria Screening Strategy Simulator

⚕ The same test sensitivity and specificity produce very different predictive values depending on background prevalence. Adjust prevalence, test characteristics, and screening strategy to observe how the surveillance pyramid changes — and how missed sub-patent infections sustain transmission.

Diagnostic Test

Test Characteristics

Sensitivity (Se) P(T+ | Disease+) 92%

Specificity (Sp) P(T− | Disease−) 97%

Sub-patent detection Sensitivity for low-density infections 10%

Epidemiological Setting

True prevalence Including sub-patent infections 10%

Sub-patent fraction Of all true infections 30%

Screening Strategy

Population coverage Fraction screened/treated 80%

Cost per treatment (ACT) USD $2.5

Operating Metrics

PPV

—

pos. pred. value

NPV

—

neg. pred. value

Missed cases

—

per 10,000

False treats

—

per 10,000

NNS

—

need to screen

Total cost

—

per 10,000 pop.

Programme Cost Breakdown

Testing—

True-positive treatment—

False-positive treatment—

Total—

Confusion Matrix

Fig. I · N = 10,000

Test Positive

Test Negative

Disease +

Disease −

True Positive

—

Detected & treated

False Negative

—

Missed — reservoir

False Positive

—

Unnecessary Rx

True Negative

—

Correctly cleared

Adjust prevalence and test characteristics to observe how the cell counts shift.

PPV & NPV across Prevalence

Fig. II · Bayesian sweep

The vertical line marks the current prevalence. PPV collapses at low prevalence even with high sensitivity — the core Bayesian insight.

Test Comparison — ROC

Fig. III

Each test's operating point on the ROC plane. PCR dominates; microscopy and RDT trade sensitivity for cost.

Missed Cases vs Prevalence

Fig. IV

Missed infections per 10,000 screened. Sub-patent reservoir is invisible to RDT and microscopy.

Surveillance Pyramid & Transmission Reservoir

Fig. V

Stacked bar showing the fate of all true infections under the current screening strategy. Missed sub-patent infections sustain onward transmission.

Field Interpretation

Adjust parameters to generate interpretation.

Bayesian Diagnostic Equations

PPV = Se·P / [Se·P + (1−Sp)·(1−P)]

NPV = Sp·(1−P) / [Sp·(1−P) + (1−Se)·P]

TP = N·P·[Se_patent·(1−f_sub) + Se_sub·f_sub]

FN = N·P − TP LR+ = Se/(1−Sp)

P=prevalence · f_sub=sub-patent fraction · Se_sub=sub-patent sensitivity

Malaria Diagnostics &Screening Strategy

Bayesian Diagnostic Equations

Malaria Diagnostics &
Screening Strategy