Our patient cohort of combined group 3 and group 4 PH had an overall 1-year mortality rate of 20.3% and 1-year hospital readmission rate of 46.9%. However, there is significant variation in clinical outcomes when causative etiologies are compared. OHS/OSA patients require frequent hospitalizations likely due to OHS exacerbations, however their overall outcome is good with low mortality.[21] This is because treatment with positive pressure and diuresis improves their underlying condition and therefore has low mortality.[22] Patients with bronchiectasis and ILD had high 1-year hospital readmissions but also had high 1-year mortality. This is likely because patients with bronchiectasis and ILD have advanced lung disease, and in the setting of PH, they have limited cardiopulmonary reserve to handle exacerbations, which would contribute to high mortality.[23, 24] However, it should be kept in mind that bronchiectasis and ILD are on a spectrum with patients having mild and others having advanced disease.[25] Therefore, we cannot generalize this to each ILD and bronchiectasis patient.
Bronchiectasis and ILD patients have a higher chance of mortality when they have concurrent PH, when compared to COPD and OHS/OSA patients. These higher risk groups also have higher mean PASP (Fig 2). This may suggest that in patients who have moderate to advanced bronchiectasis or ILD, a screening echocardiogram for PH evaluation may be considered for prognostication and management.
Diagnosing the etiology of PH, especially group 3 can be challenging, with diseases sometimes having overlapping presentations such as hypercapnic failure due to OHS or COPD. At times group 3 etiologies may have overlap syndromes such as COPD-OSA, ILD-OSA, OHS-bronchiectasis. Only 17.6% of COPD patients, and 18.4% of OHS/OSA patients had PASP elevations > 66mmHg, suggesting that these conditions are less likely to extremely high PASP. Therefore, when TTE shows severe PASP elevations, the likelihood of COPD or OHS/OSA driving those pressures may be less likely, as compared to bronchiectasis or ILD.
When comparing our data to PH-registries from other low-middle income countries (LMIC), our mortality rates were similar to the PAPUCO African Registry which had a 6-month mortality of 21%, with our cohort showing a 12-month mortality of 20.3%.[15] The PROKERALA reported an annual mortality rate of 4%, however re-hospitalization rates were greater than 50%, which is similar to our cohort at 46.9%.[17] These registries have included pediatric patients and all PH groups, and likely the inclusion of group 2 PH may decrease the overall mortality rates. A recent study done in Europe showed that patients with ILD who had borderline PH had worse outcomes than patients with COPD and borderline PH.[26] Our cohort supplements the finding that ILD patients with PH have a high 1-yr mortality of 30.4% compared to COPD patients with PH at 11.8%.
PAPUCO and PROKERALA registries have also used ECHO based diagnosis of PH, and this is a reality in LMIC, like our study. Only 3 patients (2.3%) received RHC, even though we have the expertise and equipment to perform RHC. However, cost is a major limiting factor and therefore, most physicians and patients would prefer relying on TTE as the primary diagnostic tool. This suggests that the TTE reading should be robust and standardized. Studies are suggesting good correlation of RHC with TTE findings of pulmonary artery pressures,[7, 12] however future studies could evaluate whether diagnostic utility of echocardiogram or RHC may change the diagnosis, treatment, or outcome.
On multivariate regression analysis, use of diuretics and no 1-year hospital readmission showed decreased odds of 1-year mortality (table 3). No other factors highlighted an increased odds of 1-year mortality, even though the rates of mortality were significantly different in patients with bronchiectasis compared to OHS/OSA. We suspect this is related to fewer numbers per etiology in our study, which may decrease the power and ability to find a difference on regression analysis. Prior studies have identified that high PASP, right ventricular dysfunction, and gender are correlated with increased mortality in PH.[18-20, 27] Prior studies have been conflicting regarding increased PH related mortality in men or women.[20, 27] Our study shows no increased odds of mortality based on gender, RV function or related to higher PASP pressures.
Our study has several limitations that should be kept in mind. We could not keep into account patients being readmitted to other hospitals, which suggests that our calculated 1-year readmission rate is likely underestimated. Secondly, due to the study being conduct in a LMIC, we used echocardiography as the diagnostic criteria instead of RHC. This may have overestimated the diagnosis of pulmonary hypertension as some patients with mild PH (PASP < 45 mmHg) may have lower or normal pressures on RHC. However, most of our patients (69.9%) had moderate or severe elevations in PASP, and in those situations they will likely have PH on RHC. Lastly, a portion of our patients (39.1%) did not have ideal PH work up, as only 60.9% of patients had workup to exclude pulmonary embolism.
There is a further need to study group 3 PH, especially associated treatment factors that may improve PH outcomes. In conclusion, our study indicates a high 1-year mortality and 1-year hospital readmission rate for patients with pulmonary hypertension. Patients with bronchiectasis and ILD were found to be at highest risk for worse 1-year outcomes.