Skip to main content
. 2018 Jul 23;19(1):851. doi: 10.4102/sajhivmed.v19i1.851

TABLE 5.

Diagnostic accuracy assessment of the respiratory rate and oxygen saturation clinical prediction rules for the diagnosis of Pneumocystis jirovecii pneumonia among 500 seriously ill HIV-infected participants presenting with a cough of any duration and one or more World Health Organization danger signs.

Total score Number with score Probability of PJP Sensitivity Specificity Likelihood ratio positive Likelihood ratio negative
Respiratory rate model
0 50 0 100.0% 0.0% 1.000 -
1 153 3.1% 94.6% 10.6% 1.059 0.506
2 154 6.5% 82.1% 43.5% 1.453 0.411
3 73 13.3% 69.6% 76.6% 2.973 0.396
4 45 25.2% 51.8% 90.8% 5.608 0.531
5 23 42.5% 26.8% 97.8% 11.893 0.749
6 2 61.9% 1.8% 99.8% 7.929 0.984
Oxygen saturation model
0 135 0 100 0 1.000 -
1 161 3.9% 91.1% 29.3% 1.288 0.305
2 30 7.7% 78.6% 64.0% 2.180 0.335
3 90 14.8% 76.8% 70.5% 2.603 0.329
4 41 26.4% 58.9% 88.5% 5.130 0.464
5 8 42.6% 46.4% 96.2% 12.126 0.557
6 35 60.6% 39.3% 97.1% 13.418 0.626

PJP, Pneumocystis jirovecii pneumonia.

Note: Although it is possible to score 9 points in the respiratory rate clinical prediction rule, no participant had a score of higher than 6; therefore, the probability of PJP could not be calculated for higher scores.