Factors associated with survival in patients who undergo peritoneal dialysis catheter placement following cardiac surgery

Journal of Pediatric Surgery (2013) 48, 1269–1276

www.elsevier.com/locate/jpedsurg

Factors associated with survival in patients who undergo peritoneal dialysis catheter placement following cardiac surgery☆ Arin L. Madenci a , Alexander P. Stoffan b , Satish K. Rajagopal c , Joshua J. Blinder c , Sitaram M. Emani d , Ravi R. Thiagarajan c , Christopher B. Weldon b,⁎ a

University of Michigan Medical School, Ann Arbor, MI, USA Department of Surgery, Boston Children's Hospital and Harvard Medical School, MA, USA c Department of Cardiology, Boston Children's Hospital and Harvard Medical School, MA, USA d Department of Cardiac Surgery, Boston Children's Hospital and Harvard Medical School, MA, USA b

Received 23 February 2013; accepted 8 March 2013

Key words: Peritoneal dialysis; Catheters; Intraabdominal hypertension; Cardiac surgical procedures; Acute kidney injury; Renal insufficiency

Abstract Purpose: Pediatric post-cardiac surgery patients are at risk for acute kidney injury and intraabdominal hypertension. The present study assesses indications and outcomes of postoperative peritoneal dialysis catheter (PDC) placement in this population. Methods: We retrospectively reviewed single-institution patients who underwent PDC placement postcardiac surgery between 1999 and 2011 (n = 55). Baseline, clinical course, and outcome data were recorded pre- and post-PDC. We used multivariable logistic and Cox analyses to assess factors associated with mortality. Results: In-hospital mortality of the study cohort was 67.3% (n = 37). Peritoneal dialysis was performed in 21 patients (38.2%). Five patients (9.1%) experienced adverse events related to PDC placement. Greater post-PDC decreases in abdominal girth (adjusted odds ratio [OR] = 2.43; P = 0.02) and BUN (OR = 1.06; P = 0.04) were associated with survival. Additionally, preoperative ventilator independence (hazard ratio [HR] = 1.18; P b 0.01) and lower creatinine (HR = 8.32; P b 0.01), as well as greater post-PDC decrease in inotrope score (HR = 1.33; P b 0.02) were associated with survival. Conclusions: In-hospital mortality of the study cohort was 67%. Less severe pre-PDC renal impairment, increased pre-PDC abdominal girth, and greater post-PDC improvement of abdominal girth, renal function, and inotrope requirements were associated with survival. Prospective trials are needed to assess appropriate indications and timing of PDC placement, with consideration of more aggressive treatment for intraabdominal hypertension. © 2013 Elsevier Inc. All rights reserved.

☆ Institutional review board: Approved, No. M10-11-0575. ⁎ Corresponding author. Department of Surgery, Children's Hospital Boston, Boston, MA 02115, USA. E-mail address: [email protected] (C.B. Weldon).

0022-3468/$ – see front matter © 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.jpedsurg.2013.03.022

1270 Pediatric patients who undergo cardiac surgery are at risk for multiple complications, including acute kidney injury (AKI) and intraabdominal hypertension (IAH). Both morbidities are closely linked to postoperative fluid balance disturbances as well as low cardiac output syndrome [1] and capillary leak syndrome [2,3]. In particular, renal insufficiency affects over 10% of children who undergo cardiopulmonary bypass and further aggravates fluid overload and extravasation [4–7]. Severe IAH exacerbates renal dysfunction by increasing renal venous pressure [8]. Urgent placement of a peritoneal dialysis catheter (PDC) is at times performed to decompress the abdomen and improve cardiorespiratory and renal function. In severe cases, peritoneal dialysis may be performed to temporarily replace renal function. However, the benefit of peritoneal access is unclear. The proportion of patients receiving PDCs who utilize the catheter for peritoneal dialysis is unknown. If the PDC functions merely as a conduit to drain a hypertensive abdomen, then a temporary decompressive catheter would be equally effective and less co-morbid than a tunneled, permanent PDC, because complications related to PDCs may occur in up to 20% of patients [5]. Furthermore, the suitability of a PDC to adequately decompress the abdominal cavity in cases of IAH has not been definitively proven and, as such, more aggressive intervention (i.e. decompressive laparotomy) may be warranted. However, the timing and indications for this step are unknown. The purpose of this study, therefore, is to assess the indications and outcomes of post-cardiac surgery PDC placement.

A.L. Madenci et al. children [10]. Increased abdominal girth was used as a proxy for IAH [11]. Surgical complexity was assessed for by the Risk Adjustment for Congenital Heart Surgery (RACHS-1) risk categories [12,13]. RACHS-1 categories range from 1 to 6, with 6 indicating the highest risk category. Determination of RACHS-1 risk category is primarily based on surgical case complexity and also takes into consideration patient characteristics such as age at the time of operation, prematurity, presence of major non-cardiac structural anomaly, and multiple cardiac procedures (Appendix) [12]. If a patient could be grouped into two or more RACHS-1 categories, the highest risk score was used. If a patient was unable to be assigned a RACHS-1 category (e.g. patients undergoing heart transplantation or missing data), the patient was designated “Unassigned.” Patient inotrope score was calculated based on maximal requirements over a 24-h period. Inotrope score has been validated in this population to predict postoperative morbidity such as prolonged ventilation requirement [14]. Inotrope score is determined based on the formula: [1] Total inotrope score ¼ dopamine ðμ=kg= minÞ þ dobutamineðμ=kg= minÞ þ adrenaline=epinephrineðμ=kg=minÞ  100 þ phenylephrineðμ=kg= minÞ  100 þ noradrenaline=norepinephrineðμ=kg= minÞ  100:

1.2. Outcomes

1. Methods 1.1. Patient population An institutional review board-approved, retrospective review was conducted of the medical records of patients who underwent cardiac surgery between January 1999 and December 2011 at Boston Children's Hospital. Patients who underwent postoperative permanent PDC implantation following cardiac surgery met inclusion criteria. Each cardiac surgery and PDC placement was performed by 1 of 10 and 1 of 14 experienced surgeons, respectively. Patient demographics, cardiac diagnoses, perioperative data, and outcomes were collected. Physiological variables were also recorded, including abdominal girths, BUN and serum creatinine, ventilator settings, and medication dosages. Estimated glomerular filtration rate (GFR) was calculated by the Schwartz formula and used to assess renal function prior to PDC placement, according to the pediatric-modified RIFLE criteria (pRIFLE). The RIFLE classification system was originally developed by the Acute Dialysis Quality Initiative and modified for use in children. pRIFLE strata are assigned based on incremental decreases in GFR or oligoanuria and classified as: Risk, Injury, Failure, Loss and End-stage renal disease [9]. The pRIFLE criteria have been validated for use in critically ill

The primary outcome of the study mortality. Secondary outcomes included change of renal laboratory tests, inotrope settings, and 7-day post-PDC change in (centimeters [cm]).

was in-hospital 24 h post-PDC score, ventilator abdominal girth

1.3. Statistical analysis We tabulated patient demographics, cardiac diagnoses, cardiac surgery perioperative data, PDC implantation perioperative data, and outcomes, reporting percentages based on available data. Chi-square and Fisher's exact tests, where appropriate, were used to compare categorical variables between survivors and non-survivors; continuous variables were assessed using the Mann–Whitney U test. Multivariable association with mortality was evaluated using logistic regression with stepwise elimination. Survival analysis was performed using the Kaplan–Meier method for univariable analysis of categorical variables. Additionally, multivariable Cox regression was performed using the stepwise method. Survival time was calculated as duration in days from cardiac operation to death. Patients were censored when discharged from the hospital. Patients with missing data were excluded from each univariable and multivariable analysis. A P value

Survival of patients with PDC placement after cardiac surgery b 0.05 was considered statistically significant. All analyses were conducted using SAS version 9.3 (Cary, NC).

2. Results 2.1. Baseline characteristics Fifty-five patients who underwent PDC implantation following cardiac surgery were identified. In-hospital mortality for the entire cohort was 67.3% (n = 37). Table 1

1271 Twenty-six patients were male (47.3%). There were an equivalent number of Caucasian (50.9%, n = 28) and nonCaucasian (n = 27, 49.1%) patients. Thirty-one patients (57.4%) were ventilator dependent preoperatively. The most common index cardiac surgery was the Norwood operation (21.8%, n = 12). Median RACHS-1 score was 3 (interquartile range [IQR] = 3–6), and the observed mortality rate of 67.3% exceeded the predicted mortality rate of 8.5% for patients in RACHS-1 category 3. The median age at surgery was 7 days (IQR = 4–68). Patients who survived were more likely to be Caucasian (77.8%, n = 14; vs. non-survivors, 37.8%, n = 14; P b 0.01). Overall median follow-up was

Cardiac surgery characteristics.

Variable Sample size Male Caucasian ⁎ Weight (kg) Premature birth Age at surgery (days) Neonate at surgery RACHS-1 Category 2 Category 3 Category 4 Category 5 Unassigned Cardiac diagnosis HLHS TOF TOGV AV canal Heterotaxy Other Cardiac operation Norwood I Transplant TOF repair AV canal repair Arterial switch Truncus repair Fontan Other Preoperative status Vent. dependent PaO2 SO2 BUN Creatinine Lactate Operative time (m) CCT (m) TBT (m)

Overall

Non-survivor

Survivor

55 (100.0) 26 (47.3) 28 (50.9) 3.2 (2.7–4.5) 13 (24.1) 7 (4–68) 38 (69.1)

37 (67.3) 16 (43.2) 14 (37.8) 3.1 (2.5–4.1) 8 (22.2) 8 (4–41) 26 (70.3)

18 (32.7) 10 (55.6) 14 (77.8) 3.6 (3.0–7.5) 5 (27.8) 7 (4–323) 12 (66.7)

4 19 6 12 14

(7.3) (34.5) (10.9) (21.8) (25.5)

2 9 4 10 12

(5.4) (24.3) (10.8) (27.0) (32.4)

2 10 2 2 2

(11.1) (55.6) (11.1) (11.1) (11.1)

19 6 6 4 4 16

(34.5) (10.9) (10.9) (7.3) (7.3) (29.1)

15 5 1 3 2 11

(40.5) (13.5) (2.7) (8.1) (5.4) (29.7)

4 1 5 1 2 5

(22.2) (5.6) (27.8) (5.6) (11.1) (27.8)

12 7 5 2 5 3 5 16

(21.8) (12.7) (9.1) (3.6) (9.1) (5.5) (9.1) (29.1)

10 6 4 1 1 3 1 11

(27.0) (16.2) (10.8) (2.7) (2.7) (8.1) (2.7) (29.7)

2 (11.1) 1 (5.6) 1 (5.6) 1 (5.6) 4 (22.2) 0 (0.0) 4 (22.2) 5 (27.8)

31 (57.4) 54.3 (43.7–64.1) 89.0 (83.0–94.0) 15.0 (8.0–23.0) 0.6 (0.4–0.8) 1.7 (1.1–2.6) 391 (330–526) 82 (53–118) 146 (119–250)

24 (66.7) 54.7 (45.2–66.8) 90.0 (84.0–95.0) 16.0 (9.0–30.0) 0.6 (0.5–0.9) 1.5 (1.1–2.6) 403 (342–507) 84 (53–116) 150 (119–263)

7 (38.9) 47.4 (41.3–64.1) 88.0 (75.0–91.0) 11.0 (7.0–18.0) 0.6 (0.4–0.7) 1.9 (1.6–2.5) 362 (297–526) 80 (53–123) 143 (107–203)

P 0.39 b 0.01 0.17 0.74 0.86 0.79 0.09

0.05 0.46 0.32 0.12 0.09 0.26 0.66 0.88 0.38

Data are reported as no. (%) or median (interquartile range). AV, atrioventricular; BUN, blood urea nitrogen; CCT, cross-clamp time; HLHS, hypoplastic left heart syndrome; TBT, total cardiopulmonary bypass time; TOF, tetralogy of fallot; TOGV, transposition of the great vessels; RACHS-1, risk adjustment for congenital heart surgery. ⁎ P b 0.05.

1272

A.L. Madenci et al.

67 days (IQR = 26–293 days). Details of baseline and cardiac surgery characteristics are reported in Table 1.

2.2. Impact of PDC The median time from cardiac surgery to PDC placement was 17 days (IQR = 10.0–34.0). Postoperative timing of PDC placement did not affect survival (non-survivors median, 17.0; IQR = 10.0–29.0 vs. survivors median 17.0; IQR = 11.0–38.0. P = 0.33). Twenty-four hours prior to PDC placement, 22.5% of patients were in the Risk stratum (n = 11), 18.4% were in the Injury stratum (n = 9), and 16.3% were in the Failure stratum (n = 8), as defined by the RIFLE criteria [9]. Peritoneal dialysis was performed 38.2% of patients (n = 21). All catheters were utilized as peritoneal drains after placement. Five patients (9.1%) sustained

Table 2

adverse events related to PDC placement, including bacterial peritonitis (n = 1), bile peritonitis (n = 1), periprocedural bowel perforation requiring re-exploration (n = 1), contaminated catheter requiring replacement (n = 1), and a clogged catheter (n = 1) requiring replacement. Among patients afflicted with such complications none survived (P = 0.17). Twenty-four hours prior to PDC placement, survivors were more likely than non-survivors to have a lower creatinine (median, 0.6 vs. 1.1; P b 0.01) as well as a larger abdominal girth (median, 40.0 vs. 36.0 cm; P = 0.03). Surviving patients also had greater 7-day post-PDC decreases in BUN (median, − 13.0 vs. − 3.0 mg/dL; P = 0.03) and abdominal girth (− 2.5 vs. 0.0 cm; P b 0.01). Other factors related to PDC placement are reported in Table 2. In a multivariable analysis, greater 7-day post-PDC decreases in BUN and abdominal girth were significantly associated with

PDC placement: baseline and outcomes.

Variable

Overall

Non-survivor

Sample size Time to PDC (days) Pre-PDC RIFLE score None Risk Injury Failure Pre-PDC values PS PEEP RR FiO2 BUN Creatinine ⁎ Lactate Inotrope score Abd. girth (cm) ⁎ 24-h post-PDC change a PS PEEP RR FiO2 Creatinine BUN Lactate Inotrope score Abd. girth (cm) 7-day post-PDC change b Creatinine BUN ⁎ Lactate Abd. girth (cm) ⁎ Peritoneal dialysis

55 (100.0) 17.0 (10.0–34.0)

37 (67.3) 17.0 (10.0–29.0)

21 (42.3) 11 (22.5) 9 (18.4) 8 (16.3)

12 (35.3) 9 (26.5) 6 (17.7) 7 (20.6)

Survivor 18 (32.7) 17.0 (11.0–38.0) 9 2 3 1

P 0.33 0.36

(60.0) (13.3) (20.0) (6.7)

30.0 (28.0–33.0) 5.0 (4.0–6.0) 24.0 (20.0–26.0) 0.5 (0.4–0.6) 60.5 (45.0–77.0) 0.9 (0.6–1.5) 2.2 (1.5–3.3) 10.0 (7.5–15.5) 37.5 (34.3–44.0)

30.0 (28.0–33.0) 5.0 (4.0–6.0) 22.0 (20.0–28.0) 0.5 (0.3–0.6) 62.0 (47.0–80.0) 1.1 (0.8–1.8) 2.3 (1.8–2.9) 11.5 (7.5–20.5) 36.0 (33.0–42.0)

30.5 (28.5–33.5) 5.5 (5.0–6.0) 25.0 (24.0–26.0) 0.5 (0.4–0.6) 48.0 (42.0–65.0) 0.6 (0.5–0.8) 1.6 (1.2–4.3) 8.0 (3.0–15.0) 40.0 (36.0–48.5)

0.82 0.17 0.44 0.62 0.16 b 0.01 0.40 0.09 0.03

0.0 (0.0–0.0) 0.0 (0.0–0.0) 0.0 (0.0–0.0) 0.0 (0.0–0.0) 0.0 (− 0.1–0.2) 3.0 (− 6.0–9.0) 0.0 (− 1.0–0.5) 0.0 (− 2.0–0.0) − 0.5 (− 2.0–1.0)

0.0 (0.0–0.0) 0.0 (0.0–0.0) 0.0 (0.0–0.0) 0.0 (0.0–0.0) 0.0 (− 0.2–0.2) 3.0 (− 5.0–9.0) 0.1 (− 0.8–0.5) 0.0 (− 1.0–0.0) 0.0 (− 1.5–1.0)

0.0 (− 2.0–0.0) 0.0 (0.0–0.0) 0.0 (− 2.0–0.0) 0.0 (0.0–0.0) 0.0 (− 0.1–0.1) 2.0 (− 7.0–6.0) − 0.3 (− 1.5–0.3) 0.0 (− 2.5–0.0) − 0.7 (− 2.0–0.3)

0.29 0.75 0.19 0.61 0.76 0.45 0.15 0.63 0.17

− 0.1 (− 0.4–0.1) − 5.5 (− 19.0–19.5) − 0.5 (− 1.3–0.2) − 1.0 (− 2.5–0.5) 21 (38.2)

− 0.1 (− 0.4–0.2) − 3.0 (− 11.0–29.0) − 0.5 (− 1.1–0.6) 0.0 (− 1.5–1.0) 14 (37.8)

− 0.2 (− 0.3–0.1) − 13.0 (− 21.0 to − 3.0) − 0.6 (− 2.4–0.0) − 2.5 (− 5.0 to − 1.0) 7 (38.9)

0.91 0.03 0.17 b 0.01 0.94

Data are reported as no. (%) or median (interquartile range). BUN, blood urea nitrogen; PDC, peritoneal dialysis catheter; PEEP, peak end-expiratory pressure; PS, pressure support; RIFLE; risk, injury, failure, loss, end-stage renal disease; RR, respiratory rate. a Defined as: 24-h post-PDC change in physiologic values from pre-PDC baseline (change = 24-h post-PDC − pre-PDC baseline). b Defined as: 7-day post-PDC change in physiologic values from pre-PDC baseline (change = 7 days post-PDC − pre-PDC baseline). ⁎ P b 0.05.

Survival of patients with PDC placement after cardiac surgery Table 3

Multivariable logistic regression model for mortality.

Variable

Adjusted 95% Confidence P odds ratio interval

7-day post-PDC change a BUN 1.06 Abdominal girth (cm) 2.43

1.01–1.13 1.16–5.09

0.04 0.02

BUN, blood urea nitrogen; PDC, peritoneal dialysis catheter. Stepwise backward elimination model included sex; race; weight; premature birth; age at cardiac surgery; RACHS-1 risk category N 3; preoperative laboratories (BUN, creatinine, and lactate), inotrope score, and abdominal girth; change in inotrope score [defined as: 24-h post-PDC change in physiologic values from pre-PDC baseline (change = 24-h post-PDC − pre-PDC baseline)]; change in laboratories (BUN, creatinine, and lactate) and abdominal girth [defined as: 7-day post-PDC change in physiologic values from pre-PDC baseline (change = 7 days post-PDC − pre-PDC baseline)]. a Defined as: 7-day post-PDC change in physiologic values from pre-PDC baseline (change = 7 days post-PDC − pre-PDC baseline).

survival. Adjusted odds ratios are reported in Table 3. This model had an area of 90.5% under the receiver operating characteristic curve, indicating excellent ability to accurately discriminate survival, as displayed in Fig. 1.

2.3. Survival analysis We then conducted survival analyses using the Kaplan– Meier method. Patients who were not intubated prior to cardiac surgery were more likely to survive (log-rank χ2 = 4.31, P = 0.04; Fig. 2), as were Caucasian patients (log-rank χ2 = 7.11, P b 0.01; Fig. 3). In the multivariable Cox analysis, non-ventilator dependence prior to cardiac surgery (hazard ratio [HR] = 1.18; 95% CI, 1.05–1.32; P b 0.01),

Fig. 1 Receiver operating characteristic curve for ability of multivariable model (Table 3) to discriminate survival. Area under the curve of 90.5% suggests excellent accuracy.

1273 lower pre-PDC creatinine (HR = 8.32; 95% CI, 2.46–28.16; P b 0.01), and greater 24-hour post-PDC decrease in inotrope score (HR = 1.33; 95% CI, 1.05–1.69; P = 0.02) were associated with survival (Table 4).

3. Discussion In our cohort of patients who underwent PDC implantation following cardiac surgery, in-hospital mortality was 67%. Patients with lower pre-PDC creatinine and higher prePDC abdominal girths had lower mortality. Following PDC placement, patients who showed improvement in renal function or greater decrease in abdominal girth from prePDC baseline had lower mortality. Additionally, improvement in inotrope score following PDC placement was associated with decreased mortality. The minority of PDCs were utilized for dialysis (38.2%). These results suggest that patients with more severe IAH prior to PDC placement who underwent successful abdominal decompression may have benefitted the most from the PDC. Given that dialysis was not associated with increased survival, it is likely that the mechanism of improvement for patients with IAH entailed simple decompression rather than renal replacement. No patients proceeded to undergo decompressive laparotomy as a more aggressive intervention to treat their IAH. It is unclear whether patients with poorer renal function benefitted from PDC placement. Renal function recovery following PDC placement was associated with improved survival, however the use of dialysis was not. While intraabdominal pressure was not measured directly, indirect measurement by decreasing abdominal girth suggests that relief of IAH may improve renal function. As no patients underwent a laparotomy for relief of IAH, indications for this procedure could not be assessed. Among ventilated pediatric intensive care unit patients with a baseline mortality rate of 8%, those patients with IAH who underwent decompressive laparotomy had a mortality rate of 58% [15]. Following cardiac surgery, IAH may exacerbate renal injury due to postoperative decreased GFR and worsening low cardiac output syndrome due to preload impairment [8]. Treatment of IAH may include temporary catheter placement for drainage, neuromuscular blockade, and decompressive laparotomy [15,16]. However, the precise timing, implementation, and escalation of these treatment options are not well understood. The suitability of a peritoneal drain to relieve IAH is recognized; however, there are occasions when decompressive laparotomy may be warranted [17]. PDC may alleviate postoperative AKI through several mechanisms. First, PDC placement may serve as additional source of postoperative fluid removal. Second, peritoneal dialysis removes pro-inflammatory cytokines from the circulation, thereby preventing renal damage [2,3]. For pediatric cardiac surgery patients in whom PDCs were placed perioperatively for ultrafiltration, most catheters were

1274

A.L. Madenci et al.

Fig. 2

Patients who were not ventilator dependent pre-cardiac surgery had improved survival. Log-rank χ2 = 4.31, P = 0.04.

not used for postoperative peritoneal dialysis [18]. In our study, the dialysis utilization rate was 38%. Our findings suggest that there may be two types of patients receiving PDCs, those who require abdominal decompression and those who require renal replacement therapy. For the former, early detection of IAH is crucial to prevent abdominal compartment syndrome. Though beneficial, PDC placement is not without procedural risks, including: exit-site infections, which can lead to tunnel infection and peritonitis; bowel perforation; leakage; migration; and hernia [19,20]. In pediatric patients, the rate of peritonitis is approximately one episode of peritonitis per

Fig. 3

catheter-year [20]. In our study, the complication rate was 9%, with adverse events including bacterial peritonitis, bile peritonitis, periprocedural bowel perforation, catheter infection, and obstructed catheter requiring replacement. While all patients who suffered complications died, the association did not reach statistical significance in the context of limited sample size. Furthermore, there are no definitive data to utilize in selecting which patients with IAH will benefit from a PDC and which patients may require decompressive laparotomy. Known risk factors for PDC placement include younger age, greater surgical complexity, nonelective hospital

Caucasian patients had improved survival compared with non-Caucasian patients. Log-rank χ2 = 7.11, P b 0.01.

Survival of patients with PDC placement after cardiac surgery Table 4 Variable

Multivariable Cox model for mortality. a

Hazard 95% Confidence P ratio interval

Pre-cardiac surgery 1.18 ventilator dependency Creatinine prior to PDC 8.32 Change in inotrope score a 1.33

1.05–1.32

b 0.01

2.46–28.16 1.05–1.69

b 0.01 0.02

PDC, peritoneal dialysis catheter. Stepwise backward elimination model included sex; race; weight; premature birth; age at cardiac surgery; RACHS-1 risk category N 3; preoperative laboratories (BUN, creatinine, and lactate), inotrope score, and abdominal girth; change in inotrope score a; change in laboratories (BUN, creatinine, and lactate) and abdominal girth [defined as: 7-day post-PDC change in physiologic values from pre-PDC baseline (change = 7 days post-PDC − pre-PDC baseline)]. a Defined as: 24-h post-PDC change in physiologic values from prePDC baseline (change = 24-h post-PDC − pre-PDC baseline).

admission, cardiopulmonary bypass, and acute renal failure [21]. Currently, there remains no universally accepted threshold in postoperative timing or clinical status for PDC placement. Different providers treating children with a range of diseases, complexities, and perioperative courses renders quantification and determination of the optimal time for PDC placement difficult to compute. With the present study, we have attempted to identify an at-risk population of children who may benefit from earlier intervention with PDC, potentially abrogating the deleterious secondary physiological insults from perioperative resuscitation and treatment. Patients who continue to worsen with a PDC in place are certainly candidates for decompressive laparotomy, but this measure may arrive too late once its need is recognized [15]. We were unable to compare outcomes of PDC to those of no intervention or alternative intervention, such as decompressive laparotomy. The populations receiving each intervention are different such that, without a randomized study, their comparison presents difficulties. Despite this, however, certainly fewer patients require decompressive laparotomy compared to those needing PDCs. Subjecting children with borderline circulations to a highly invasive decompressive laparotomy would undoubtedly result in significant risks without a clearly beneficial outcome. For the latter set of patients with impaired renal function, the benefits of PDC placement as a late therapy are not well understood. We did not find a significant survival benefit in patients receiving dialysis following PDC placement. Additionally, we could not assess the optimal timing of PDC placement. In one small uncontrolled study, non-survivors were shown to have a longer interval between onset of acute kidney injury and initiation of peritoneal dialysis as compared with survivors [22]. There is also considerable variability among surgeons in timing of PDC placement for cardiac surgery patients [23]. In the present study, timing did not affect survival. However, we were unable to control for confounding by

1275 indication. Prospective randomized and controlled trials will be crucial to evaluate the benefits and optimal timing of PDC placement in postoperative patients with poor renal function. We found that Caucasian race was associated with increased survival (P b 0.01). This effect persisted, when adjusting for covariates of prematurity and RACHS-1 risk category. Our study was not designed to assess reasons for racial disparities in post-cardiac surgery outcomes. However, this finding is in accordance with the literature, in which mortality rates were higher in African American patients compared with Caucasian patients [24], even when accounting for demographic, clinical, and access to care covariates [25]. Hypotheses for this continued disparity include referral patterns and timing of cardiac surgery [25], biological differences [26], and even provider bias which remains welldocumented among adults [27].

3.1. Limitations There were several limitations to our findings. As a retrospective study, there was increased potential for bias by confounding. We used multivariable analyses when possible to address this possibility. Additionally, our sample size was relatively small and may have underpowered our statistical analysis. Furthermore, this was a single-center study at an institution that does not routinely place PDCs. Confounding by indication as well as the preferences of individual providers may have led to earlier or later PDC placement and other differences in clinical management. We expect that these differences in clinical decision making would be nondifferential and bias the results toward the null. Temporary drain placement was unable to be assessed, because it was not adequately recorded in the clinical records. As such, we were unable to make a direct comparison between PDCs and temporary drains. Likewise, we were unable to assess the utility of prophylactic PDCs due to an exceedingly small sample size. Secondly, we used laboratory and physical exam measurements as proxies for AKI and IAH, respectively. In particular, for IAH, abdominal girth may have limited sensitivity compared to bladder pressure measurement [28]. However, tracking change in abdominal girth over time with respect to PDC placement provided a useful indicator and was the best measure available. We also incorporated changes in ventilator settings and inotrope requirements to assess the possible physiological impact of alleviating IAH. Along the same lines, we were unable to evaluate with certainty the clinical indication for PDC placement, which could only be approximated by change in markers of renal function or IAH. In conclusion, the patients who undergo PDC implantation following cardiac surgery are a high-risk population, with an in-hospital mortality of 67%. Patients with higher pre-PDC abdominal girths and greater post-PDC decreases in abdominal girth and inotrope requirement were more likely to survive. Permanent peritoneal access may not be necessary

1276 for patients with IAH. Prospective randomized and controlled trials are needed to assess the appropriate indications and timing of PDC placement in patients following cardiac surgery, with consideration of more aggressive treatment for intraabdominal hypertension.

A.L. Madenci et al.

[10]

[11]

Appendix Examples of surgeries corresponding to RACHS-1 risk classification categories. RACHS-1 class Example of surgery 1 Atrial septal defect repair, Aortopexy 2 Pulmonary valve replacement, Total repair of tetralogy of fallot 3 Aortic valve replacement, Repair of tetralogy of Fallot with pulmonary atresia 4 Repair of truncus arteriosus, Repair of hypoplastic or interrupted aortic arch 5 Repair of truncus arteriosus and interrupted arch 6 Norwood operation Unassigned Heart transplant Adapted from Jenkins et al. [12]. RACHS-1, Risk Adjustment in Congenital Heart Surgery.

References [1] Hoffman TM, Wernovsky G, Atz AM, et al. Efficacy and safety of milrinone in preventing low cardiac output syndrome in infants and children after corrective surgery for congenital heart disease. Circulation 2003;107(7):996-1002. [2] Bokesch PM, Kapural MB, Mossad EB, et al. Do peritoneal catheters remove pro-inflammatory cytokines after cardiopulmonary bypass in neonates? Ann Thorac Surg 2000;70(2):639-43. [3] Dittrich S, Aktuerk D, Seitz S, et al. Effects of ultrafiltration and peritoneal dialysis on proinflammatory cytokines during cardiopulmonary bypass surgery in newborns and infants. Eur J Cardiothorac Surg 2004;25(6):935-40. [4] Kist-van Holthe tot Echten JE, Goedvolk CA, Doornaar MB, et al. Acute renal insufficiency and renal replacement therapy after pediatric cardiopulmonary bypass surgery. Pediatr Cardiol 2001;22(4):321-6. [5] Pedersen KR, Hjortdal VE, Christensen S, et al. Clinical outcome in children with acute renal failure treated with peritoneal dialysis after surgery for congenital heart disease. Kidney Int Suppl 2008;108: S81-6. [6] Stromberg D, Fraser Jr CD, Sorof JM, et al. Peritoneal dialysis. An adjunct to pediatric postcardiotomy fluid management. Tex Heart Inst J 1997;24(4):269-77. [7] Chiravuri SD, Riegger LQ, Christensen R, et al. Factors associated with acute kidney injury or failure in children undergoing cardiopulmonary bypass: a case-controlled study. Paediatr Anaesth 2011;21(8):880-6. [8] Doty JM, Saggi BH, Blocher CR, et al. Effects of increased renal parenchymal pressure on renal function. J Trauma 2000;48(5):874-7. [9] Bellomo R, Ronco C, Kellum JA, et al. Acute renal failure— definition, outcome measures, animal models, fluid therapy and

[12]

[13]

[14]

[15]

[16]

[17]

[18]

[19]

[20] [21]

[22]

[23]

[24]

[25]

[26]

[27]

[28]

information technology needs: the Second International Consensus Conference of the Acute Dialysis Quality Initiative (ADQI) Group. Crit Care 2004;8(4):R204-12. Akcan-Arikan A, Zappitelli M, Loftis LL, et al. Modified RIFLE criteria in critically ill children with acute kidney injury. Kidney Int 2007;71(10):1028-35. Patel A, Lall CG, Jennings SG, et al. Abdominal compartment syndrome. AJR Am J Roentgenol 2007;189(5):1037-43. Jenkins KJ, Gauvreau K, Newburger JW, et al. Consensus-based method for risk adjustment for surgery for congenital heart disease. J Thorac Cardiovasc Surg 2002;123(1):110-8. Boethig D, Jenkins KJ, Hecker H, et al. The RACHS-1 risk categories reflect mortality and length of hospital stay in a large German pediatric cardiac surgery population. Eur J Cardiothorac Surg 2004; 26(1):12-7. Davidson J, Tong S, Hancock H, et al. Prospective validation of the vasoactive-inotropic score and correlation to short-term outcomes in neonates and infants after cardiothoracic surgery. Intensive Care Med 2012;38(7):1184-90. Pearson EG, Rollins MD, Vogler SA, et al. Decompressive laparotomy for abdominal compartment syndrome in children: before it is too late. J Pediatr Surg 2010;45(6):1324-9. De Wolf A, Poelaert J, Herck I, et al. Surgical decompression for abdominal compartment syndrome after emergency cardiac surgery. Ann Thorac Surg 2008;85(6):2133-5. Cheatham ML, Malbrain ML, Kirkpatrick A, et al. Results from the International Conference of Experts on Intra-abdominal Hypertension and Abdominal Compartment Syndrome II. Recommendations. Intensive Care Med 2007;33(6):951-62. Alkan T, Akcevin A, Turkoglu H, et al. Postoperative prophylactic peritoneal dialysis in neonates and infants after complex congenital cardiac surgery. Asaio J 2006;52(6):693-7. Werner HA, Wensley DF, Lirenman DS, et al. Peritoneal dialysis in children after cardiopulmonary bypass. J Thorac Cardiovasc Surg 1997;113(1):64-8 [discussion 68–70]. Ronnholm KA, Holmberg C. Peritoneal dialysis in infants. Pediatr Nephrol 2006;21(6):751-6. Madenci AL, Thiagarajan RR, Stoffan AP, et al. Characterizing peritoneal dialysis catheter use in pediatric patients after cardiac surgery. J Thorac Cardiovasc Surg. Forthcoming 2012. Chien JC, Hwang BT, Weng ZC, et al. Peritoneal dialysis in infants and children after open heart surgery. Pediatr Neonatol 2009;50(6): 275-9. Sorof JM, Stromberg D, Brewer ED, et al. Early initiation of peritoneal dialysis after surgical repair of congenital heart disease. Pediatr Nephrol 1999;13(8):641-5. Sistino JJ, Ellis Jr C. Effects of health disparities on survival after neonatal heart surgery: why should racial, ethnic, gender, and socioeconomic status be included in the risk analysis? J Extra Corpor Technol 2011;43(4):232-5. Oster ME, Strickland MJ, Mahle WT. Racial and ethnic disparities in post-operative mortality following congenital heart surgery. J Pediatr 2011;159(2):222-6. Mahle WT, Kanter KR, Vincent RN. Disparities in outcome for black patients after pediatric heart transplantation. J Pediatr 2005;147(6): 739-43. Schulman KA, Berlin JA, Harless W, et al. The effect of race and sex on physicians' recommendations for cardiac catheterization. N Engl J Med 1999;340(8):618-26. Malbrain ML. Different techniques to measure intra-abdominal pressure (IAP): time for a critical re-appraisal. Intensive Care Med 2004;30(3): 357-71.