Characteristics of NVA and VA-HARTI
The univariate analysis showed that NVA- and VA-HARTI patients were similar in terms of patients’ demographics, admitting diagnosis, surgeries, and complications, although their ICU courses and severity were different (Table 1). When comparing five groups, there were significant differences among them, most notably in their ICU course and complications (Table S3). We also found that patient characteristics remained mostly stable during ten years while multiple aspects of medical and surgical practice have changed. Most notably, the use of mechanical ventilation, benzodiazepines, urinary catheters, and antibiotics decreased significantly (Table S4).
Table 1
Baseline clinical characteristics of NVA-HARTI and VA-HARTI patients; for baseline characteristics of other patient groups refer to Table S3. The p-values are obtained from t-test test for numerical variables and Chi-square test for categorical variables. Bonferroni-Holm correction for multiple comparisons was used to adjust p-values. Yellow highlighting indicates characteristics with p-value below the level of significance. Abbreviations: CI, confidence interval; CNS, central nervous system; CSFL-NE, CSF leakage from nose and/or ears; CSFL-SS, CSF leakage from surgical site; EVD, external ventricular drain; EETS, Endoscopic Endonasal Trans-sphenoidal Surgery; HAI, healthcare-associated infections; ICU, intensive care unit; ICP, intracranial pressure; INSD, Implantation or Revision of a Neurosurgical Device; NVA-HARTI, non-ventilator-associated healthcare-associated respiratory tract infections; PBSS, Pittsburgh Brain Stem Score; RASS, Richmond Agitation-Sedation Scale, SSSI, superficial surgical site infections; VA-HARTI, ventilator-associated healthcare-associated respiratory tract infections.
|
NVA-HARTI (n = 254)
|
VA-HARTI (n = 577)
|
p-value
|
Adjusted p-value
|
No of patients (%) or median
|
95% CI or 1st and 3rd quartiles
|
No of patients (%) or median
|
95% CI or 1st and 3rd quartiles
|
Demographics
|
Age, years
|
52.0
|
[31.25; 62.0]
|
47.0
|
[31.0; 60.0]
|
0.06101
|
0.88955
|
Male patients
|
153 (60.2%)
|
[54.2; 66.3]
|
310 (53.7%)
|
[49.7; 57.8]
|
0.09597
|
0.96763
|
Charlson Comorbidity Index
|
3.0
|
[2.0; 5.0]
|
3.0
|
[2.0; 5.0]
|
0.10293
|
0.96763
|
Admitting diagnosis
|
Patients with traumatic brain injuries
|
58 (22.8%)
|
[17.7; 28.0]
|
142 (24.6%)
|
[21.1; 28.1]
|
0.64303
|
0.99987
|
Patients with brain tumors
|
127 (50.0%)
|
[43.9; 56.1]
|
278 (48.2%)
|
[44.1; 52.3]
|
0.68315
|
0.99987
|
Patients with neurovascular disorders
|
59 (23.2%)
|
[18.0; 28.4]
|
142 (24.6%)
|
[21.1; 28.1]
|
0.73342
|
0.99987
|
Patients with other neurosurgical disorders
|
10 (3.9%)
|
[1.5; 6.3]
|
15 (2.6%)
|
[1.3; 3.9]
|
0.41261
|
0.99975
|
Surgeries
|
Patients undergoing INSD
|
80 (31.5%)
|
[25.8; 37.2]
|
253 (43.8%)
|
[39.8; 47.9]
|
0.00107
|
0.04925
|
Patients undergoing non-cranial surgeries
|
119 (46.9%)
|
[40.7; 53.0]
|
341 (59.1%)
|
[55.1; 63.1]
|
0.00139
|
0.06076
|
Patients undergoing craniotomy
|
172 (67.7%)
|
[62.0; 73.5]
|
402 (69.7%)
|
[65.9; 73.4]
|
0.6312
|
0.99987
|
Patients undergoing endovascular neurosurgical procedures
|
26 (10.2%)
|
[6.5; 14.0]
|
70 (12.1%)
|
[9.5; 14.8]
|
0.50304
|
0.99975
|
Patients undergoing EETS
|
9 (3.5%)
|
[1.3; 5.8]
|
24 (4.2%)
|
[2.5; 5.8]
|
0.82104
|
0.99987
|
Patients undergoing spinal surgery
|
5 (2.0%)
|
[0.3; 3.7]
|
17 (2.9%)
|
[1.6; 4.3]
|
0.56576
|
0.99976
|
The total number of all surgeries during one ICU admission
|
2.0
|
[1.0; 3.0]
|
2.0
|
[1.0; 3.0]
|
< 0.0001
|
0.00173
|
The total duration of all craniotomies performed during one admission, minutes
|
300.0
|
[240.0; 370.0]
|
320.0
|
[240.0; 440.0]
|
0.0982
|
0.96763
|
The total duration of all INSDs performed during one admission, minutes
|
160.0
|
[90.0; 200.0]
|
150.0
|
[90.0; 247.5]
|
0.77106
|
0.99987
|
The total duration of all EETSs performed during one admission, minutes
|
270.0
|
[215.0; 341.25]
|
230.0
|
[115.0; 315.0]
|
0.2878
|
0.99943
|
The total duration of all endovascular neurosurgical procedures performed during one admission, minutes
|
135.0
|
[90.0; 240.0]
|
140.0
|
[90.0; 260.0]
|
0.92609
|
0.99987
|
The total duration of all non-cranial surgeries performed during one admission, minutes
|
192.5
|
[101.25; 267.5]
|
187.5
|
[150.0; 278.75]
|
0.43105
|
0.99975
|
The total duration of all spinal surgeries performed during one admission, minutes
|
345.0
|
[307.5; 390.0]
|
240.0
|
[215.0; 392.5]
|
0.35991
|
0.99975
|
The total duration of all surgeries during one ICU admission
|
318.5
|
[226.0; 435.5]
|
346.0
|
[224.0; 497.0]
|
0.11787
|
0.97678
|
Severity of patient condition during ICU stay
|
Glasgo Coma Scale
|
10.0
|
[7.0; 12.0]
|
7.0
|
[7.0; 10.0]
|
< 0.0001
|
< 0.0001
|
RASS
|
-2.0
|
[-2.0; 1.0]
|
-2.0
|
[-2.0; 1.0]
|
0.00226
|
0.09459
|
PBSS
|
11.0
|
[10.0; 11.0]
|
11.0
|
[10.0; 11.0]
|
< 0.0001
|
< 0.0001
|
ICU course
|
Days with antibiotics
|
14.0
|
[8.0; 27.0]
|
24.0
|
[15.0; 42.0]
|
< 0.0001
|
< 0.0001
|
Days with anxiolytics
|
0.0
|
[0.0; 2.0]
|
0.0
|
[0.0; 1.0]
|
0.33964
|
0.99975
|
Days with arterial line
|
0.0
|
[0.0; 6.0]
|
6.0
|
[0.0; 11.0]
|
< 0.0001
|
< 0.0001
|
Days with central line
|
11.0
|
[7.0; 21.0]
|
19.0
|
[13.0; 32.0]
|
< 0.0001
|
< 0.0001
|
Days with EVD
|
0.0
|
[0.0; 2.0]
|
0.0
|
[0.0; 7.0]
|
0.01494
|
0.43557
|
Days with feeding tube
|
12.0
|
[7.0; 25.5]
|
23.0
|
[14.0; 41.0]
|
< 0.0001
|
< 0.0001
|
Patients with hypothermia
|
2 (0.8%)
|
[0.0; 1.9]
|
37 (6.4%)
|
[4.4; 8.4]
|
0.0008
|
0.03751
|
Patients with ICP monitoring
|
54 (21.3%)
|
[16.2; 26.3]
|
228 (39.5%)
|
[35.5; 43.5]
|
< 0.0001
|
< 0.0001
|
Patients with mechanical ventilation
|
229 (90.2%)
|
[86.5; 93.8]
|
577 (100.0%)
|
[100.0; 100.0]
|
< 0.0001
|
< 0.0001
|
Patients with mutism
|
40 (15.7%)
|
[11.3; 20.2]
|
144 (25.0%)
|
[21.4; 28.5]
|
0.00431
|
0.16943
|
Patients with persistent vegetative state
|
33 (13.0%)
|
[8.9; 17.1]
|
154 (26.7%)
|
[23.1; 30.3]
|
< 0.0001
|
0.00109
|
Patients with sedation
|
142 (55.9%)
|
[49.8; 62.0]
|
395 (68.5%)
|
[64.7; 72.2]
|
0.00066
|
0.03162
|
Patients with seizures
|
22 (8.7%)
|
[5.2; 12.1]
|
49 (8.5%)
|
[6.2; 10.8]
|
0.9567
|
0.99987
|
Patients with total parenteral feeding
|
14 (5.5%)
|
[2.7; 8.3]
|
45 (7.8%)
|
[5.6; 10.0]
|
0.30017
|
0.99944
|
Patients with urinary catheter
|
238 (93.7%)
|
[90.7; 96.7]
|
569 (98.6%)
|
[97.7; 99.6]
|
0.00024
|
0.01225
|
Patients with vasopressors
|
125 (49.2%)
|
[43.1; 55.4]
|
399 (69.2%)
|
[65.4; 72.9]
|
< 0.0001
|
< 0.0001
|
Complications
|
Days with intestinal dysfunction
|
0.0
|
[0.0; 4.0]
|
3.0
|
[0.0; 9.0]
|
< 0.0001
|
< 0.0001
|
Days with urinary tract infection
|
0.0
|
[0.0; 3.0]
|
2.0
|
[0.0; 14.0]
|
< 0.0001
|
< 0.0001
|
Patients with bloodstream infection
|
24 (9.4%)
|
[5.9; 13.0]
|
76 (13.2%)
|
[10.4; 15.9]
|
0.16037
|
0.99108
|
Patients with CNS infection
|
37 (14.6%)
|
[10.2; 18.9]
|
100 (17.3%)
|
[14.2; 20.4]
|
0.37464
|
0.99975
|
Patients with CSFL-NE
|
42 (16.5%)
|
[12.0; 21.1]
|
122 (21.1%)
|
[17.8; 24.5]
|
0.14899
|
0.98908
|
Patients with CSFL-SS
|
19 (7.5%)
|
[4.2; 10.7]
|
62 (10.7%)
|
[8.2; 13.3]
|
0.1819
|
0.99339
|
Patients with intestinal dysfunction
|
120 (47.2%)
|
[41.1; 53.4]
|
363 (62.9%)
|
[59.0; 66.9]
|
< 0.0001
|
0.00183
|
Patients with other infections
|
43 (16.9%)
|
[12.3; 21.5]
|
144 (25.0%)
|
[21.4; 28.5]
|
0.01379
|
0.41816
|
Patients with SSSI
|
11 (4.3%)
|
[1.8; 6.8]
|
54 (9.4%)
|
[7.0; 11.7]
|
0.01895
|
0.50729
|
Patients with urinary tract infection
|
83 (32.7%)
|
[26.9; 38.4]
|
320 (55.5%)
|
[51.4; 59.5]
|
< 0.0001
|
< 0.0001
|
Of the 3,937 ICU admissions, 859 had at least one episode of HARTI with a total of 955 discrete HARTI cases, group distribution shown at Fig. 1. Recurrent HARTI cases were observed in 79 ICU admissions.
Among patients who had a single HARTI event during their hospital stay, NVA-HARTI occurred on median ICU day 4, with the onset of the half of the cases between ICU days 3 and 7 (Figure S1A). The onset of VA-HARTI occurred later, on median ICU day 7 [Q1,Q3: 5, 12] and on median day 6 of mechanical ventilation [Q1,Q3: 4, 9], Figures S1A, S1B. The symptoms of NVA-HARTI were present for a median of 7 days [Q1,Q3: 3, 11], as compared to 9 days [Q1,Q3: 5, 16] for VA-HARTI (Figure S1C).
Prevalence of HARTI
The mean daily prevalence of NVA-HARTI was 5.6% and there was no longitudinal time trend observed, p-value = 0.85 (Figure S2A). The mean prevalence of VA-HARTI was 13.7% and it significantly decreased from 2011 to 2020, p-value = 0.00011 (Figure S2B). For both infections, the prevalence fluctuated significantly from day to day, and the extracted annual trend had irregular periodic patterns.
Incidence of HARTI
The mean cumulative incidence of NVA-HARTI was 7.2 (6.4-8.0) per 100 ICU admissions that did not change significantly over time (p-value = 0.72). The mean cumulative incidence of VA-HARTI was 15.4 (14.2–16.5) per 100 ICU admissions; it significantly reduced from 19.0 (14.6–23.4) in 2011 to 11.7 (8.7–14.6) in 2020, p-value = 0.008 (Fig. 2A).
The mean incidence rate of NVA-HARTI was 4.2 ± 2.0 per 1000 patient-days in the ICU with no time trend, p-value = 0.15. The mean incidence rate of VA-HARTI was 9.5 ± 3.0 per 1000 patient-days, also with no time trend, p-value = 0.09, Fig. 2B.
When normalized to ventilator-days, VA-HARTI incidence rate didn’t change throughout the study period (p-value = 0.79) and averaged at 16.4 ± 4.7 cases per 1000 ventilator-days (Fig. 2C). At the same time, the total number of ventilator-days per year decreased by 30%, from 3,690 in 2011 to 2,565 in 2020.
Length of Stay
NVA-HARTI group had a significantly higher median hospital LOS (32 days [21; 48.5]) comparing to No HAI (17 days [11; 25]), though it was significantly shorter than in the VA-HARTI group, 46 days [28; 76.5], Fig. 3A. From 2011 to 2020, hospital LOS did not change in the NVA-HARTI group although it decreased in all other groups. Additionally, the variance of hospital LOS decreased in all groups (Figure S3A).
The ICU LOS in the NVA-HARTI group was 15 days [10; 28.75], significantly higher than in No HAI and Other HAI groups. The longest ICU LOS was in the VA-HARTI group with a median of 26 days [17; 43]; the shortest ICU LOS was in No HAI group with a median of 5 days [3; 7], Fig. 3B. Over time, the ICU LOS remained unchanged for both HARTI groups but decreased significantly for all other groups (Figure S3B).
Mortality Rates
The mean all-cause in-hospital mortality in the study population was 10.8% (9.8–11.8) over ten years. It declined significantly from 16.4% (12.2–21) in 2011 to 5.4% (3.1–7.9) in 2020, p-value = 0.0001 (Fig. 3E).
In the NVA-HARTI group, the mortality rate was 12.3% (7.9–16.8), and no different from other groups. In VA-HARTI patients, the mortality rate was 16.7% (13.6–19.7), twice that of the No HAI group, 8.2% (7.0-9.4), p-value < 0.0001, Fig. 4C. From 2011 to 2020, the mortality rate dropped significantly in all groups except for the NVA-HARTI, Figure S3C.
Survival Analysis and the Impact of HARTI on Mortality
Patients with NVA-HARTI, VA-HARTI, and with no HAI died on median hospital days 35 [12.5; 57], 35.5 [18; 75.25], and 10 [6; 18] respectively, Figure S4A. This difference represented the immortal time bias. To account for it, we adjusted for time-independent variables (age, sex, diagnosis, and CCI) in a Cox time varying model. Neither infection was found to be independently associated with mortality, NVA-HARTI p-value = 0.13 (Table S5A), VA-HARTI p-value = 0.82 (Table S5B).
Survival curves were obtained from the competing risk model while accounting for discharge as a competing event, Fig. 3D. The Kaplan-Meier model underestimated the risk of death and was deemed inappropriate (Figure S4B). The survival probability difference function showed that during the first 20 days in the hospital, patients in the No HAI group had a significantly higher chance of dying than patients in both HARTI groups. Later on, survival curves crossed swapping the risk between groups: No HAI and VA-HARTI at approximately hospital day 30 and No HAI and NVA-HARTI at approximately hospital day 45. After hospital day 52, patients in both HARTI groups had higher probability of death and this equilibrium continued up to hospital day 228 when the last patient in the No HAI group was discharged or died (Fig. 3D).
In stepwise logistic regression, VA-HARTI was an independent risk factor of death with an odds ratio of 1.54 (1.11–2.14), p-value = 0.009 (Table 2A, all covariates in Table S6A). NVA-HARTI was not a predictor of mortality, odds ratio 1.43 (0.99–2.06), p-value = 0.057 (Table 2B, all covariates in Table S6B). The performance testing in five-fold cross-validation yielded a ROC-AUC score of 0.63 for NVA-HARTI and 0.79 for VA-HARTI models.
Table 2A
Multivariate logistic regression evaluating mortality in all study patients; the estimate for VA-HARTI. Only significant covariates are shown, for full data table refer to Table S6A. Abbreviations: CSFL-NE, CSF leakage from nose and/or ears; ICU, intensive care unit; SSSI, superficial surgical site infections; VA-HARTI, ventilator-associated healthcare-associated respiratory tract infections.
Factor
|
Odds ratio
|
Lower band of 95% CI
|
Upper band of 95% CI
|
p-value
|
Days with bloodstream infection
|
1.07
|
1.01
|
1.12
|
0.0136
|
Days with intestinal dysfunction
|
1.03
|
1.01
|
1.05
|
0.0041
|
Days with SSSI
|
1.02
|
1.00
|
1.03
|
0.0087
|
Days with tracheostomy
|
0.96
|
0.94
|
0.98
|
0.0001
|
Days with urinary catheter
|
1.02
|
1.00
|
1.05
|
0.0430
|
Days with vasopressors
|
1.02
|
1.00
|
1.04
|
0.0379
|
Glasgo Coma Scale
|
0.59
|
0.57
|
0.62
|
< 0.0001
|
Patients undergoing craniotomy
|
0.73
|
0.59
|
0.88
|
0.0015
|
Patients with brain tumors
|
1.86
|
1.38
|
2.51
|
< 0.0001
|
Patients with CSFL-NE
|
1.95
|
1.08
|
3.50
|
0.0259
|
Patients with persistent vegetative state
|
0.27
|
0.16
|
0.44
|
< 0.0001
|
Patients with traumatic brain injuries
|
0.44
|
0.26
|
0.75
|
0.0024
|
The total number of all surgeries during one ICU admission
|
1.20
|
1.06
|
1.36
|
0.0039
|
VA-HARTI
|
1.54
|
1.11
|
2.14
|
0.0090
|
Table 2B
Multivariate logistic regression evaluating mortality in all study patients; the estimate for NVA-HARTI. Only significant covariates are shown, for full data table refer to Table S6B. Abbreviations: CSFL-NE, CSF leakage from nose and/or ears.
Factor
|
Odds ratio
|
Lower band of 95% CI
|
Upper band of 95% CI
|
p-value
|
Days with antibiotics
|
0.97
|
0.94
|
0.99
|
0.0072
|
Days with central line
|
1.02
|
1.00
|
1.04
|
0.0323
|
Days with intestinal dysfunction
|
1.03
|
1.01
|
1.05
|
0.0011
|
Days with urinary catheter
|
1.03
|
1.01
|
1.06
|
0.0014
|
Days with vasopressors
|
1.04
|
1.02
|
1.06
|
<0.0001
|
Patients with brain tumors
|
1.75
|
1.36
|
2.25
|
<0.0001
|
Patients with CSFL-NE
|
1.86
|
1.32
|
2.62
|
0.0004
|
Risk of HARTI
The risk of both HARTI types increased rapidly in the beginning of ICU stay. The NVA-HARTI probability grew faster and plateaued earlier than VA-HARTI, around day 8 in the ICU with the highest daily risk of 1.0% at that time. Whereas the risk of VA-HARTI continued to increase up to ICU day 20 before reaching a plateau phase (Fig. 4A). Also, a secondary peak of daily infection risk occurred earlier for NVA-HARTI, Fig. 4B. During the first week in the ICU, the risk of infection doubled every 18.4 hours for NVA-HARTI and every 13.6 hours for VA-HARTI.
The cumulative probability of VA-HARTI depending on the number of days on mechanical ventilation increased rapidly during the first 8 ventilation days, Fig. 4C. The probability of VA-HARTI doubled daily during the first week on ventilation. A secondary peak became apparent after ventilation day 36 (Fig. 4D).