Study population and clinical data
A retrospective analysis was performed of all patients who had received non-pharmacological intervention (renal artery angioplasty with or without stenting, unilateral nephrectomy) for the treatment of renovascular hypertension in a defined geographical area in the West of Scotland (population 2.4 million) across 3 centres between 2010 and 2022.
A database of patients with mechanical intervention performed for hypertension was used and augmented by a word-search of the radiology information system in which all imaging and interventions are reported. Case-notes were reviewed, and data obtained.
Data extracted included patient demographics as well as the indication for intervention and the systolic and diastolic blood pressure measurements before intervention and at the point of discharge following intervention. The number of prescribed antihypertensive medications was recorded at each of these time periods. 2 patients who died in the immediate perioperative period were excluded from this analysis.
Previously published data
For comparative analysis, published data from two multi-centre studies of renal artery stenting were accessed10,11. Measurements relating to blood pressure and medication load before and after intervention in each study were collected to facilitate calculation of the HTi in each study cohort.
The Hypertensive Index
A novel marker of excessive blood pressure and medication load was empirically derived based on the systolic blood pressure and number of antihypertensive drugs.
$${\text{Hypertensive}}\;{\text{Index}} = \left( {{\text{Systolic}}\;{\text{Blood}}\;{\text{Pressure}} – {12}0} \right) \times \left( {{\text{No}}.\;{\text{Antihypertensives}} + {1}} \right)$$
The number of antihypertensive medications was supplemented by a factor of 1 to allow calculation of the hypertensive index in those requiring no antihypertensive medications.
A subgroup of patients who did not achieve significant blood pressure change was then assessed using the Hypertensive Index to evaluate for a more sensitive measure of treatment effect.
Comparison with other measures
Receiver operator characteristic (ROC) analysis was used to compare the HTi against conventional measurements (systolic or diastolic blood pressure alone; medication load alone).
Various iterations of the HTi formula were tested to assess changes in the sensitivity and specificity depending on the blood pressure measurement component; including mean arterial pressure, diastolic pressure and pulse pressure.
To compare the Hypertensive Index against the GWAS method, pre- and post-intervention scores were calculated for each patient and the reduction in these scores compared. Scores pertaining to the GWAS method were calculated by the addition of 10 mmHg to the patients’ systolic blood pressure if they were prescribed any antihypertensive drug. ROC analysis was used to compare the efficacy of each in determining whether intervention had been ‘successful’.
Further analysis was performed to assess the potential utility of the pre-intervention HTi in predicting the impact of treatment. The published data from two multi-centre studies of renal artery stenting were analysed10,11. Mean pre- and post-stenting HTi scores calculated for each cohort, and their outcomes compared.
Statistical analysis
‘Significant change’ in blood pressure was defined as a reduction in systolic blood pressure of ≥ 20 mmHg, or diastolic blood pressure of ≥ 10 mmHg12.
For the purposes of ROC analysis, ‘successful intervention’ was defined by a significant change in blood pressure without additional antihypertensive medications, or by any reduction in medication load without an increase in blood pressure.
Data was collated using Microsoft Excel (Version 16.53 © Microsoft 2021). Statistical analysis was performed using RStudio (Version 1.4.1717© 2009–2021 RStudio, PBC). Data distribution was assessed using Shapiro-Wilks Test. Means were compared by Paired T-test, Welch Two-Sample T-test or Mann–Whitney U Test as appropriate.
Ethics
All participants provided written informed consent and all study procedures were in concordance with the Declaration of Helsinki, 1975 (revised 2000). Formal ethical approval was not required as per national legislation produced by the Health Research Authority13.