There has been a dramatic increase in the number of adults with CHD in recent times, due to multiple factors including but not limited to improvements in early diagnosis, improved surgical techniques, and post-operative care.Reference Warnes1,Reference Moodie2 In developed countries, the majority of children with CHD survive into adulthood.Reference Brida and Gatzoulis3 However, in developing countries, the story is different. There is a huge burden of adult patients with unrepaired CHD.Reference Shahabuddin, Hashmi, Rakhshan, Khan, Sami and Amanullah4 This older age at operation may be associated with longer ICU and hospital stay.Reference Gnanappa, Ganigara and Prabhu5,Reference Mello, Carvalho, Baucia and Magalhaes Filho6
Prolonged length of stay has been identified as a significant adverse outcome after adult CHD surgery. There are several factors – pre-, intra-, and post-operative – that contribute to prolonged length of stay after CHD surgery. These include patient age,Reference Kempny, Dimopoulos and Uebing7 pre-existing comorbid disease, prolonged cardiopulmonary bypass time,Reference Kogon, Grudziak and Sahu8 and post-operative complications, such as infection or acute kidney injury.Reference Cedars, Benjamin, Burns, Novak and Amin9 Moreover, length of stay is considered as an important indicator for hospital resource consumptionReference Romano, Hussey and Ritley10 and hospital quality of care.Reference Thomas, Guire and Horvat11 It is also associated with significantly increased cost for the patients,Reference Kalish, Daley, Duncan, Davis, Coffman and Iezzoni12 which can be a major issue in the developing countries where resources are meagre.
Though there have been numerous studies exploring the factors associated with prolonged length of stay after surgery for CHD in the paediatric population,Reference al-Haddad, Menk, Kochilas and Vinocur13,Reference Mercer-Rosa, Elci and DeCost14 there has been very little published on the adult population undergoing surgery for CHD. Data are particularly scarce for developing countries, such as Pakistan. These challenges contribute to a different spectrum of clinical, surgical, and post-operative care for adult CHD patients.
Thus, this study explores the factors that are associated with prolonged post-operative length of hospital stay in adults undergoing surgery for CHD at a tertiary-care hospital in Pakistan.
Materials and methods
Study location and sample
This retrospective cross-sectional study was conducted at the Aga Khan University Hospital. All adult patients (≥18 years) who underwent cardiac surgery with cardiopulmonary bypass for their CHD from January 2011 to December 2016 at Aga Khan University Hospital were included in this study. Non-probability consecutive sampling was used for all patients fulfilling the inclusion criteria. Patients undergoing cardiac surgery for acquired cardiac disease were not included in this study.
Material and procedure
Data were collected from medical records using the online patient data system through a structured pro forma. This included patients’ demographics, pre-operative variables such as primary diagnosis, saturations, body mass index,15 pre-operative left ventricular function, haemoglobin level, associated comorbid conditions (such as diabetes, obesity, hypertension, arrhythmias, etc.), and previous surgeries. Patients were classified by the procedures they underwent into the risk categories described in the Risk Adjustment for the Congenital Heart Surgery-1 system.Reference Jenkins, Gauvreau, Newburger, Spray, Moller and Iezzoni16
Intraoperative details included the type of CHD repair (univentricular versus biventricular repair), cardiac surgery performed, cardiopulmonary bypass time, aortic cross-clamp time, intraoperative inotrope score, and significant perioperative blood loss. Perioperative bleeding was categorised according to the Universal Definition for Perioperative Bleeding classification.Reference Dyke, Aronson and Dietrich17
Post-operative parameters included post-operative serum creatinine, hypotension on-arrival to cardiac ICU (CICU), maximum inotropic score, occurrence of post-operative complications (such as acute kidney injury, bleeding, infection, arrhythmias, low cardiac output, drain output, need for reopening and death, etc.), duration of mechanical ventilation, duration of CICU stay, and total length of stay. Acute kidney injury was defined by the criteria recommended by the Kidney Disease Improving Global Outcomes in 2012.18
Prolonged length of stay was defined as a post-operative hospital stay of longer than the 75th percentile of the overall cohort, which in our study was defined as, which is the cutoff used by previous studies as well.Reference Krell, Girotti and Dimick19 The 75th percentile for hospital length of stay was 8 days, and hence, the prolonged length of stay was defined as a length of hospital stay >8 days.
Statistical analysis
Data were analysed using IBM Statistical Package for Social Sciences Version 21.0. Descriptive statistics were presented as mean ± standard deviation and n (%) for continuous and categorical variables. Independent sample t-test and chi-square test were used for the comparison of mean and percentages, respectively. Univariate and multivariable logistic regression models were used to establish the relationship between the exposures and outcomes. Variables were included in multivariable model on the basis of clinical likelihood and significance (p < 0.25) when analysed in univariate model. Odds ratios, 95% confidence interval, and p-values are presented in tables. A p-value of <0.05 was considered as significant.
Results
Demographics and pre-operative characteristics
A total of 166 patients fit the inclusion criteria for this study, the mean age being 32.05 (±12.11) years. According to the Risk Adjustment for Congenital Heart Surgery-1, 86 (51.8%) of the CHD diagnoses fell in Category 1, 50 (30.1%) fell in Category 2, and 30 (18.1%) fell in Category 3. Comorbid disease was seen in 98 (59.0%) patients.
A total of 38 patients (22.9%) experienced prolonged length of stay. The majority of the patients with prolonged length of stay were males (71.1%), and their mean body mass index was significantly lower than patients without a prolonged length of stay (p = 0.001). There was a significantly greater percentage of underweight patients among those with prolonged length of stay (44.7 versus 23.4%; p = 0.010) and a significantly greater percentage of obese patients in the group without prolonged length of stay (41.4 versus 21.1%; p = 0.022). The mean haemoglobin in patients with prolonged length of stay was significantly higher as compared to those without prolonged length of stay (p = 0.006). However, the higher anaemia could indicate the existence of cyanotic disease, as patients with cyanosis had a significantly higher haemoglobin than those without (17.6 ± 4.19 versus 13.0 ± 1.84 g/dl; p < 0.001). While the majority of patients had normal left ventricular function pre-operatively (83.7%), 19 (11.5%) had mild left ventricular dysfunction, and 8 (4.8%) had moderate left ventricular dysfunction.
Risk Adjustment for Congenital Heart Surgery-1 Category 1 formed 61.7% of the group without prolonged length of stay as compared to only 18.4% of the group with prolonged length of stay (p < 0.001). Common comorbids included obesity (36.7%), hypertension (14.5%), and anaemia (6.0%). Infective endocarditis was seen in 7.9% of patients who had prolonged length of stay as compared to in none among patients without a prolonged length of stay (p = 0.011). Additionally, 28.3% of patients in our study were underweight. Table 1 shows the demographics and pre-operative characteristics of the cohort.
Table 1. Demographics and pre-operative characteristics
BMI = body mass index; BSA = body surface area; cr. = creatinine; pre-op. = pre-operative; RACHS-1 = Risk Adjustment for Congenital Heart Surgery-1.
Intraoperative and post-operative characteristics
The majority (42.2%) of operations performed were for atrial septal defect closure. The type of cardiac operation performed showed a significant association with prolonged length of stay. The most common operation in patients without prolonged length of stay was atrial septal defect closures (50.0%) as compared to tetralogy of Fallot repair (18.4%) and aortic valve repair (18.4%) among patients with prolonged length of stay (p = 0.005). Operations included in the Others category were mitral valve repair, pulmonary valve replacement, partially anomalous pulmonary venous connection surgery, right ventricular infundibulectomy, Ebstein anomaly correction, and Fontan operation.
The mean cardiopulmonary bypass time and aortic cross-clamp time were both significantly higher in patients with prolonged length of stay (both p < 0.001). Moreover, 28.9% of patients with prolonged length of stay suffered perioperative excess blood loss as compared to 10.9% in patients without prolonged length of stay (p = 0.005). The mean inotrope score was also significantly higher in group with prolonged length of stay (p = 0.009).
Post-operatively, the duration of mechanical ventilation and the length of stay in the CICU were significantly longer in patients with prolonged length of stay (both p < 0.05). The post-operative creatinine clearance was significantly lower in patients with prolonged length of stay (p < 0.001). Table 2 shows intraoperative and post-operative characteristics of the patients.
Table 2. Intraoperative and post-operative characteristics
ASD = atrial septal defect; AV = aortic valve; CPBT = cardiopulmonary bypass time; IO = intraoperative; PO = post-operative; TOF = tetralogy of Fallot; VSD = ventricular septal defect; XCT = aortic cross-clamp time.
When analysed according to the type of surgery performed, mean length of stay was shortest in patients undergoing atrial septal defect repair (6.73 ± 2.19 days) and aortic valve repair (7.89 ± 1.63 days) and longest in tetralogy of Fallot repair (9.83 ± 6.79 days) and other surgeries (10.0 ± 8.75 days).
Post-operative complications and outcomes
Post-operative complications were seen in a significantly higher percentage of patients with prolonged length of stay as compared to those without prolonged length of stay (57.9 versus 32.8%; p = 0.005). Moreover, acute kidney injury development was also significantly associated with prolonged length of stay, with 60.5% of patients with prolonged length of stay developing kidney injury as compared to 20.3% without prolonged length of stay (p < 0.001). However, acute kidney injury in most patients was Stage 1 (67.3%), and all patients had normalisation of creatinine by discharge and at the follow-up. Patients with acute kidney injury were more likely to have had excess perioperative blood loss (29 versus 9%; p = 0.002) and also had a higher incidence of post-operative hypotension (26.3 versus 15.0%; p = 0.011) as compared to those who did not develop acute kidney injury.
Surgical site bleeding (13.2 versus 3.1%; p = 0.030), post-operative pneumonia (7.9 versus 0.8%; p = 0.038), and post-operative pleural effusion (10.5 versus 0.8%; p = 0.010) were significantly more common in patients with prolonged length of stay. The incidence of post-operative arrhythmia in the population was 7.8%.
Patients who were underweight had a significantly higher incidence of any post-operative complication (53.2 versus 32.8%; p = 0.015) and post-operative pleural effusion (8.5 versus 0.8%; p = 0.023) as compared to patients not underweight.
Univariate regression
On univariate regression, male gender, Risk Adjustment for Congenital Heart Surgery-1 Category 2 and Category 3, pre-operative haemoglobin, cardiopulmonary bypass time, aortic cross-clamp time, perioperative excess blood loss, intraoperative inotrope score, post-operative day 1 creatinine, post-operative CICU stay duration, any post-operative complication, and development of acute kidney injury were independently positively associated with prolonged length of stay. Body mass index and atrial septal defect repair were negatively associated with prolonged length of stay. The results of the univariate analysis are shown in Table 3.
Table 3. Univariate regression analysis
p < 0.25 included in multivariable analysis.
Multivariable regression
All variables analysed in the univariate analysis with p < 0.25 were included in the multivariable analysis. Results of the multivariable analysis showed that intraoperative aortic cross-clamp time and post-operative acute kidney injury were independently associated with prolonged length of stay. Moreover, the results also showed that body mass index and ventricular septal defect repair were independently negatively associated with prolonged length of stay. Out of these variables, post-operative acute kidney injury development showed the strongest association (odds ratio: 7.761; 95% confidence interval: 1.029–58.510). The results of the multivariable analysis are shown in Table 4.
Table 4. Predictors of prolonged length of stay – multivariable analysis
Discussion
This study aimed to identify the factors associated with prolonged length of stay in adult patients undergoing surgery for CHD. This would allow hospitals to identify patients at increased risk for having prolonged length of stay and concentrate efforts toward the prevention of prolonged length of stay and early patient discharge. Our study finds that factors associated with prolonged length of stay include lower body mass index, longer aortic cross-clamp time, and development of post-operative acute kidney injury.
The mean patient age of 32.05 years in our study is comparable to others by Kwiatkowski et al (median: 29 years).Reference Kwiatkowski, Price and Axelrod20 Most patients belonged to Risk Adjustment for Congenital Heart Surgery-1 Category 1 (51.8%) and Category 2 (30.1%) in our study. This is contrast to studies by Kogon et al (Category 1: 12%, Category 2: 39%, and Category 3: 42%),Reference Kogon, Grudziak and Sahu8 van Gameren et al (Category 1: 29.7%, Category 2: 26.9%, and Category 3: 42.8%),Reference van Gameren, Putman, Takkenberg and Bogers21 and Bhatt et al (Category 3: 74.4%).Reference Bhatt, Rajabali, He and Benavidez22 In developing countries such as Pakistan, a lack of access to appropriate care for pediatric CHD leads to a mortality rate as high as 60% among children.Reference Bode-Thomas23 Those with milder defects tend to survive, often undiagnosed, into adulthood, leading to the majority of adult congenital heart disease (ACHD) defects being Risk Adjustment for Congenital Heart Surgery-1 Category 1 or Category 2.Reference Shahabuddin, Hashmi, Rakhshan, Khan, Sami and Amanullah4,Reference Saxena24
Acute kidney injury occurred in 29.5% of our total patients and was present in 60.5% of patients with prolonged length of stay and was strongly associated with prolonged length of stay on multivariable analysis. Acute kidney injury contributes significant post-operative morbidity, hence contributing to prolonged length of stay.Reference Mariscalco, Lorusso, Dominici, Renzulli and Sala25–Reference Cedars, Benjamin, Burns, Novak and Amin27 Moreover, acute kidney injury development is also associated with longer duration of ventilation and ICU stay post-operatively.Reference Fuhrman, Nguyen, Sanchez-de-Toledo, Priyanka and Kellum28,Reference Memon, Akhtar, Martins, Ahmed, Saeed and Shaheen29 Studies exploring the predictors of post-operative acute kidney injury development have found several associated factors. These include pre-operative left ventricular dysfunction, age ≥35 years, pre-operative creatinine clearance, previous arrhythmias, Risk Adjustment for Congenital Heart Surgery-1 Categories 2 and 3, aortic cross-clamp time, cardiopulmonary bypass time, excess intraoperative blood loss, and post-operative use of vancomycin.Reference Kwiatkowski, Price and Axelrod20,Reference Memon, Akhtar, Martins, Ahmed, Saeed and Shaheen29
Our results show that a lower body mass index significantly increases the likelihood of prolonged length of stay. While this is contradictory to previous studies performed,Reference Peitz, Troyer and Jones30,Reference Almashrafi, Alsabti, Mukaddirov, Balan and Aylin31 it should be noted that most patients in our study had a body mass index in the normal range of 18.5–24.9 kg/m2 which is lower than the body mass index of patients in previous studies. Moreover, our study showed a significantly greater percentage of underweight patients in those having prolonged length of stay and a significantly greater percentage of obese patients in those without prolonged length of stay. Additionally, the incidence of post-operative complications, such as pleural effusion, was also higher in underweight patients. Thus, our results favour the notion that underweight patients are at increased risk for longer length of stay due to post-operative complications, which has been shown in cardiac surgery.Reference Potapov, Loebe and Anker32 Previous studies show both obesity and underweight to confer greater risk for post-operative morbidity,Reference Lui, Saidi and Bhatt33,Reference O’Byrne, Kim and Hornik34 with obesity being specifically associated with longer length of stay.Reference Buelow, Earing and Hill35 However, a study by Brida et al in 2015Reference Brida, Dimopoulos and Kempny36 reported a protective effect of obesity in adult CHD patients, showing that further research on the association of body mass index with outcomes post-surgery for adult CHD is needed. Nevertheless, appropriate pre-operative nutritional assessment and optimisation are warranted in surgery for ACHD, especially in developing countries where malnutrition is common.
A longer aortic cross-clamp time was independently associated with a higher likelihood for prolonged length of stay in our study. Longer aortic cross-clamp time has been found to be an independent predictor of post-operative mortality and morbidity in several clinical studies,Reference Doenst, Borger, Weisel, Yau, Maganti and Rao37–Reference Al-Sarraf, Thalib and Hughes39 including post-operative acute kidney injury development.Reference Memon, Akhtar, Martins, Ahmed, Saeed and Shaheen29 Moreover, longer aortic cross-clamp time is also associated with longer hospital length of stay in cardiac surgery.Reference Cislaghi, Munari, Corona and Condemi40,Reference Al-Sarraf, Thalib and Hughes41
On multivariable analysis, ventricular septal defect repair (Category 2) was associated with a lower likelihood of prolonged length of stay. This is consistent with previous studies that show that a higher Risk Adjustment for Congenital Heart Surgery-1 category is associated with longer length of ICUReference Larsen, Pedersen, Jacobsen, Johnsen, Hansen and Hjortdal42 and hospital stay.Reference Boethig, Jenkins, Hecker, Thies and Breymann43 Risk Adjustment for Congenital Heart Surgery-1 category itself was not associated with prolonged length of stay on multivariable analysis. This is possibly due to our patient population being adults with CHD, whereas previous studies reported the association with longer length of stay in paediatric patients. Moreover, since none of the patients in our study belonged to Risk Adjustment for Congenital Heart Surgery-1 categories 4–6, we were unable to adequately explore the association between Risk Adjustment for Congenital Heart Surgery-1 categories and prolonged length of stay.
Limitations
There were several limitations to this study. The sample size used was relatively small due the limited volume of adult patients in the institution who were operated for adult CHD. Moreover, this study was a retrospective analysis of patient files that presented within the institution. All the patients belonged to Categories 1–3. Patients with more complicated surgeries belonging to Categories 4–6 were not assessed, and thus, our findings may not be generalisable to the higher Risk Adjustment for Congenital Heart Surgery-1 Categories.
Conclusion
Our study finds that predictors of prolonged length of stay include lower body mass index, longer aortic cross-clamp time, and development of post-operative acute kidney injury. Thus, pre-operative, intra-operative, and post-operative factors may contribute to prolonged length of stay in adult patients undergoing CHD surgery. Shorter operations, improved pre-operative nutritional optimisation, and timely management of post-operative complications could help prevent prolonged length of stay. Identification of these factors would help to reduce prolonged stay and associated costs. Further studies with larger cohorts involving patients with higher Risk Adjustment for Congenital Heart Surgery-1 categories and increased comorbidities are needed.
Acknowledgement
We acknowledge the Pediatric Cardiology Department and Cardiothoracic Surgery Department of Aga Khan University Hospital Karachi for granting us unlimited access to their surgical and echocardiography database. We also acknowledge the Research and Development Wing of The Society for Promoting Innovation in Education for providing mentorship to author Z.S.D. (Medical student) on this project. Society for Promoting Innovation in Education is actively involved with innovation, education, and research in the academic and public health sector.
Financial Support
This research received no specific grant from any funding agency, commercial, or not-for-profit sectors.
Conflict of interest
None.
Ethical standards
The study conducted was a retrospective study with approval from The Ethics Review committee in The Aga Khan University. The study was conducted in accordance to the ethical standards described in the 1964 Declaration of Helsinki and its later amendments.
