| |
Discussions
of the overall specificity and sensitivity of unidimensional dyspnea
scores for determining the likelihood of disease are markedly
limited. We are beginning to explore the sensitivity and specificity
of dyspnea questionnaires as well as the predictive values of
patient selections of descriptors. The sensitivity of the best,
second best, and third best phrase that seems to capture the quality
of the breathing discomfort in patients with COPD (0.93), interstitial
lung disease (ILD) (0.93), or asthma (0.85) is high (Schwartzstein
and Harver, unpublished data). On the other hand, specificity
is low (0.16, 0.29, 0.31, respectively). The positive predictive
value of "chest tightness" is 0.07 for COPD; 0.00 for
interstitial lung disease; and 0.93 for asthma. The negative predictive
value of "effort and work of breathing" for COPD, ILD,
and asthma is 0.75, 0.81, and 0.49, respectively. Although preliminary,
these data suggest that patient selections of descriptors are
sensitive, demonstrate positive and negative predictive value,
and may help determine competing diagnoses, or explanations, for
complaints of dyspnea.
|