43 patients with end stage renal disease (Ga) who have started renal replacement therapy (RRT) were included in the study. 25 were males and 18 were females (mean age 63 , range 29-81).17 were on chronic haemodialysis and 17 were on chronic ambulatory peritoneal dialysis. The mean number of months on RRT was 19, range 1-98. 16 patients suffered from diabetes mellitus (DM) and hypertension (HT), 12 patients had HT but no DM, 5 patients had DM but no HT, and 10 patients had neither DM nor HT.

48 healthy normal volunteers (Gc1) were included for comparison of data. 28 were males and 20 were females (mean age 32, range 22-57). 28 patients with diabetes mellitus and/or hypertension, serum creatinine levels <1.13mg/dL (100µmol/l) and with no evidence of diabetic and/or hypertensive nephropathy formed group Gc2. 18 of these were males and 10 were females (mean age 63, range 40-78). Patients in the latter group were matched with patients in Ga for age, gender, DM, HT and side –dominant or nondominant- were the Doppler examination was carried out.

Patients who had already failed bilateral wrist arteriovenous fistulas, had valvular heart disease, uncompensated heart failure, atrial fibrillation or were on Beta blockers were excluded from the study. Current smokers or who stopped smoking one month or less from index examination were also excluded. Informed consent was obtained from all patients and subjects. The study was approved by the local research ethics committee.

A continuous wave Doppler ultrasound machine (Bidop ES-100V3 ®) with an attached 8MHz probe was used for the study. It was calibrated once at the beginning of the study. Its accuracy is up to ±10% or less. All the examinations were carried out by a single author who had a month period of training in the procedure before the commencement of the study.

Continuous wave Doppler examination of the radial artery before and after cough was performed on all patients and subjects. The maximal change in waveform after cough was recorded. Where possible, the non-dominant upper extremity (ND) was used for the examination. If patients had failed or had functioning wrist arteriovenous fistula in the ND, the dominant upper extremity was used for the examination.

Patients were in the sitting position with the forearm to be examined resting on the thigh. The Doppler probe was held at a 60o angle to the skin. The systolic velocity (cm/sec), mean flow velocity (cm/sec), diastolic velocity (cm/sec), resistance index (systolic velocity minus diastolic velocity divided by systolic velocity), pulsatility index (peak to peak change divided by mean flow velocity), the pulse rate (bpm) were recorded three times in successive manner from the radial artery at the wrist near the level of the radial styloid process at the point of maximal pulsation. The average of the three sets of readings was used for data analysis.

Patients were then asked to cough without moving their hands The same parameters were recorded at the point of maximal change in waveform after cough. This was repeated after another two more coughs, each separated by one minute. The average of the three sets of readings was used for data analysis.

The independent student t-test for equal and unequal variances was used for analysis of data using SPSS version 12.0. Statistical significance was assumed if the P value was <0.05.

Before cough, subjects in Gc1 had significantly higher systolic velocity (S), diastolic velocity (D) and mean flow velocity (MV) than group Ga as shown in table 1. However, patients in Ga had a significantly higher pulsatility index (PI) and a higher resistance index (RI) than Gc1 (Table 1).

Figure 2

Table 1: Baseline systolic velocity (S) / cm sec-1, diastolic velocity (D) / cm sec-1, mean flow velocity (MV) / cm sec-1, pulsatility index (PI) and resistance index (RI) in Ga and Gc1. Data are means ± standard deviation.

There was no significant difference in baseline S, D, PI and RI between Gc2 and their controls from Ga as shown in table 2. However, patients in Gc2 had a significantly higher MV than their controls from Ga (table 2).

Figure 3

Table 2: Baseline systolic velocity (S) / cm sec-1, diastolic velocity (D) / cm sec-1, mean flow velocity (MV) / cm sec-1, pulsatility index (PI) and resistance index (RI) in Ga (matched patients) and Gc2. Data are means ± standard deviation.

After cough there was a significant difference in the decrease in S in Ga compared to Gc1 (0.55 ± 3.81 vs. 5.03 ± 6.03, P <0.001, 95% CI 2.40, 6.56). Also, the decrease in D was much less in Ga than Gc1 after cough (0.23 ± 0.99 vs. 3.72 ± 3.57, P <0.001, 95% CI 2.41, 4.56). After cough, MV decreased much less in Ga than in Gc1 (0.68 ± 1.20 vs. 4.65 ± 4.53, P <0.001, 95% CI 2.61, 5.33) (Figure 2).

Figure 4

Figure 2: Decrease in systolic velocity (S) / cm sec-1, diastolic velocity (D) / cm sec-1 and mean flow velocity (MV) / cm sec-1 after cough in Ga and Gc2.

There was a significant difference in the increase in PI after cough as a ratio of the baseline in Ga when compared with Gc1 (-0.36 ± 0.56 vs. -1.93 ± 2.09, P <0.001, 95% CI 0.94, 2.19) (Figure 3).

Figure 5

Figure 3: Change in Pulsatility Index (PI) after cough as a ratio of the baseline in Ga and Gc2.

Also, there was a significant difference in the increase in RI after cough as a ratio of the baseline in Ga when compared with Gc1 (-0.02 ± 0.07 vs. –0.21 ± 0.19, P <0.001, 95% CI 1.36, 2.51) (Figure 4).

Figure 6

Figure 4: Change in Resistance Index (RI) after cough as a ratio of the baseline in Ga and Gc2.

There was no significant difference in the change in S (P=0.09, 95% CI 0.03, 4.14), D (P=0.06, 95% CI 0.03, 1.39), MV (P=0.06, 95% CI 0.03, 2.00), PI as a ratio of the baseline (P=0.07, 95% CI 0.07, 1.52) and RI as a ratio of the baseline (P=0.08, 95% CI 0, 0.10) between Gc2 and their controls from Ga (data not shown).