Patient group characteristics

Over three years, our unit admitted 126 patients born at less than 30 weeks gestation, 26 of whom had moderate to severe BPD. Table 1 lists the demographic characteristics and the general therapies applied to the BPD patient group. Patients were not given antihypertensive medications other than those listed in the table.

Table 1 Characteristics of moderate to severe BPD patients and BPD patients with three or more or less than three hypertensive daysMeasurement numbers

Covering the first 90 days of life, our database contained 19,481 measured blood pressure data separately for systolic blood pressure (SBP), diastolic blood pressure (DBP), and mean blood pressure (MBP). Therefore, the average number of raw data points per patient was 749.3 for each blood pressure category. This amounted to an average of 8.3 daily blood pressure measurements per patient. For the investigated period, the median of the measurement numbers was 180 (IQR 318.5–125).

The evolution of the average systolic, diastolic and mean blood pressures in the BPD patient group compared to the reference blood pressure data

We calculated the daily average systolic, diastolic and mean blood pressure values of the BPD patient group. Based on the daily average blood pressure data, we generated two trendlines: one for the BPD group and the other for the reference data. We compared the blood pressure curve and the corresponding trendline to the reference group's average daily blood pressure. The BPD patients' SBP trendline initially had lower values and rose gradually until around the 70th day of life, when it crossed the reference trendline (see Fig. 1a). The DBP and the MBP showed similar trends to that of the SBP, but they reached the reference trendline around the 30th and the 45th days of life, respectively (see Figs. 1b and c). We performed a separate statistical analysis for the time intervals before and after the intersection of the trendlines. The calculations revealed a significant correlation between the BPD-related data and the reference data (SBP: R1 = 0.732, R2 = 0.754, p < 0.001; DBP: R1 = 0.454, R2 = 0.625, p < 0.005; MBP: R1 = 0.371, R2 = 0.699, p < 0.012). However, there was a significant difference between the paired SBP, DBP and MBP curves (the calculated averages of the differences were as follows: SBP: M1 = -1.681, p < 0.001, M2 = -0.542, p = 0.139; DBP: M1 = -1.235, M2 = 1.627, p < 0.001; and MBP: M1 = -1.454, M2 = 1.502, p < 0.001).

Fig. 1

Systolic (a), diastolic (b) and mean (c) blood pressure curves and linear trendlines of the BPD group and the reference group as a function of time. SBP, systolic blood pressure; BPD, bronchopulmonary dysplasia; ref, reference group; DBP, diastolic blood pressure

Comparison of the daily average blood pressure to the 95th-percentile blood pressure trendline

We calculated and compared each patient's daily average blood pressure to the 95th-percentile trendline. The average number of daily blood pressure measurements per patient was 8.3. At least three daily measurements were taken for all patients, ensuring that a single higher measurement was not used to diagnose hypertension. The average daily measurements per patient during the hypertensive episodes were 9.2 and 9.4 for the SBP and DBP, respectively.

Considering the 2340 patient days, we found 38 systolic (1.6%) and 50 diastolic (2.1%) daily average measurements above the 95th percentile. In our whole sample, only nine patients had no daily average blood pressure spikes (i.e., a systolic or diastolic daily average blood pressure above the 95th percentile), and 65% (n = 17) had at least one day with a hypertensive average daily blood pressure in the first three months of life. We found 13 patients (50%) who had average daily SBP and 16 patients (61.5%) who had average daily DBP values that exceeded the 95th percentile. Most of them had occasional daily blood pressure spikes. Eleven patients (42%) had three or more days with an average blood pressure above the 95th percentile.

We compared the patients with elevated blood pressure values for three or more days to the remaining patients (see Table 1). Our detailed statistical analysis showed that patients with three or more hypertensive days had a 25% chance of having AKI (H0: TP = 0.25, exact sig. = 0.034). AKI episodes (91.6%) mainly occurred within the first two weeks of life. All the investigated BPD patients had normal kidney function (i.e. age-appropriate serum creatinine levels and urine output) after the fifth week of life.

We also examined the onset of the blood pressure increase. Most daily average hypertensive blood pressure measurements occurred at the 2nd, 9th, 12th, and 13th weeks of life, corresponding to the 28th, 35th, 38th, and 39th week of corrected gestational age, respectively (see Fig. 2). We found that 34%, 22%, and 44% of the hypertensive daily average blood pressures occurred in the first, second, and third months of life, respectively.

Fig. 2

Number of systolic, diastolic, and total blood pressure spikes over 13 weeks. BP, blood pressure; Total, either systolic or diastolic BP spikes

We investigated fluid bolus, blood transfusion and inotropic data in the database and found that the duration of administration did not coincide with periods of high blood pressure. Inotropic support was administered to 73% of the patients. All patients received blood transfusions, and the average transfusion number was 7.6 per patient for 90 days. A fluid bolus was administered to 16 patients. The average number of treatments administered was 1.75 per patient. These therapies were given mainly within the first three weeks of life: 44% of blood transfusions, 88% of fluid boluses and 62% of inotropes. The use of these therapies was limited compared to the entire investigated time interval. Therefore, these factors should not influence the results of our blood pressure analysis.

Comparison of the weekly average blood pressure to the 95th-percentile blood pressure trendline

We calculated the weekly average blood pressure values for every BPD patient. The weekly average SBP and DBP values were above the 95th percentile in 3 (11.5%) of the 26 patients for one week only. These results indicate that some of the shorter hypertensive periods may remain hidden when weekly averages or averages from longer time intervals are used to obtain blood pressure measurements.