This study found a statistically significant association between the development of moderate to severe BPD and Ureaplasma spp. colonisation in ELGANs. Multiple logistic regression analyses of demographic and obstetric variables indicated that airway colonisation with U. parvum and U. urealyticum is independently associated with the development of moderate or severe BPD in ELGANs (OR 4.32; 95% CI: 1.20–15.49). Unlike similar studies that mostly screened infants for Ureaplasma spp. colonisation soon after birth, this study included infants that had their microbiological samples taken later during treatment due to a clinical suspicion of an infection or developing BPD. Thus, almost all of the infants (186; 95%) included in this study developed BPD according to the NHLBI diagnostic criteria. We analysed the differences between the group without or with mild BPD and the group with moderate or severe BPD, as there is an important clinical and prognostic distinction between the two groups [21].
Our findings are concordant with the findings of meta-analyses from Schelonka et al. and Lowe et al. that reported a statistically significant association between Ureaplasma spp. colonisation and the development of BPD36 [15, 22]. BPD36 corresponds to moderate and severe BPD according to the NHLBI criteria. On the contrary, Zheng et al. performed a meta-analysis of studies that only looked at colonisation with U. urealyticum, and did not find an association with BPD development [23]. This study did not find a difference in BPD development between U. parvum and U. urealyticum colonised infants, however U. parvum colonisation was 3-times as frequent (36 infants) compared to U. urealyticum colonisation (12 infants). Glaser et al. have similarly found U. parvum in 30/40 infants included in their study and U. urealyticum in 7/40 infants, 3/40 infants were colonised with both [19].
Recent research in the field has focused on additional risk factors for BPD development that could act synergistically with Ureaplasma spp. colonisation. Inatomi et al. have shown that, after controlling for other risk factors, Ureaplasma spp. positive infants were not at increased risk for the development of moderate to severe BPD. However, the association between the presence of Ureaplasma spp. and the risk for moderate/severe BPD increased significantly in infants on mechanical ventilation ≥ 2 weeks [24]. Similarly, Glaser et al. found an increased risk for the development of BPD in Ureaplasma-colonised infants that were mechanically ventilated for 5 days or more, but the colonisation itself was not associated with a higher risk of developing BPD [19].
In this study, retrospective data analysis showed the increasing incidence of U. parvum and U. urealyticum colonisation among ELGANs in Slovenia's largest tertiary perinatal centre in the period 2009–2019, and a lower average incidence compared to other studies. After reviewing other studies, Viscardi and Hasday in 2009 concluded that respiratory tract colonisation with Ureaplasma spp. in very low birth weight infants (< 1500g) is 25–48%, using PCR as the method of detection [3]. In 2011, Sung et al. reported 65% of newborns < 26 weeks GA were Ureaplasma spp. colonised in their study [25]. In a more recent trial from 2020, Viscardi et al reported 44/121 (36%) infants < 29 weeks GA were Ureaplasma spp. positive at one or more time points [17]. In this retrospective study, the average incidence of respiratory tract colonisation with Ureaplasma spp. among ELGANs was only 8.8%.
A possible explanation for the lower incidence of Ureaplasma spp. colonisation in this study is a slightly different population of tested infants and the relatively late time of sample collection. Aforementioned studies included and collected microbiological samples from all GA < 26 or < 29 weeks infants aged < 72 hours. In this study, the microbiological samples were collected later during treatment and only from infants that were clinically suspicious for developing BPD or signs of infection. The average time of testing for Ureaplasma spp. in our institution was the 23th day after birth. Payne et al. found in their prospective study in 2011 the highest detection rate of U. parvum in aspirates of ELGANs 3–5 days after birth [26]. It is worth noting however, that different patterns of Ureaplasma spp. colonisation have been reported. Castro-Alcaraz et al. discuss three patterns of U. urealyticum colonisation of preterm infants: persistent, early transient, and late transient. In their study, only the persistently positive colonisation pattern was associated with a significantly increased risk of developing BPD [27]. It is thus possible that screening all ELGANs soon after birth would have rendered a higher incidence of colonisation in our institution as well. Lastly, the incidence of Ureaplasma spp. colonisation among ELGANs and/or women of childbearing age could actually be lower in our geographical area, but we could not find any studies that investigated this.
Data also showed an increasing frequency of testing for Ureaplasma spp. during 2009–2019 in our institution. In 2019, there were 15% fewer newborn ELGANs than in 2009. However, there were 4.4 times more collected microbiological samples for Ureaplasma spp. in 2019 compared to 2009. This could in part explain the increasing incidence of colonisation. Another contributing factor to the increasing incidence could be the improvement of microbiological analysis, as the PCR method of detection which has a higher sensitivity than culture has been introduced in the third year of this 11-year study [28].
The strength of this study is that we captured the data of a relatively big cohort of infants in a long, 11-year period. We included infants treated in Slovenia's largest tertiary perinatal centre where 80% or more of all ELGANs receive intensive care. An experienced team of doctors and nurses has been working according to the same criteria and standards for many years, so the year-by-year results could be compared with greater reliability.
This study has several limitations. Firstly, because the study was performed retrospectively, we were not able to control the exact indications for or the timing of Ureaplasma spp. testing and other variables. The decision to collect microbiological samples was in the hands of the treating physician. Another limitation was the NHLBI diagnostic criteria itself, as we could not include infants that died before the 28th day of age. Additionally, we could not use the recently proposed new BPD severity diagnostic criteria that includes roentgenological imaging, as it was not routinely performed in our institution [29].
In conclusion, this study found an increasing incidence rate of Ureaplasma spp. airway colonisation in ELGANs treated in our institution in the period 2009–2019. This could in part be due to a greater proportion of tested infants and greater sensitivity of microbiological methods. We also discovered a statistically significant correlation between the severity of BPD and the colonization with Ureaplasma spp. Controlling for other risk factors for BPD in the regression model, the preterm infants colonized with Ureaplasma spp. at the average age of 23 days had 4.32-times higher odds for the development of moderate or severe BPD. Screening of all ELGANs soon after birth and early treatment of Ureaplasma spp. colonisation could lead to a reduction of the incidence of BPD.