Abstract
Introduction:
Extended hospitalization of low birth weight infants increases risk of medical and psychosocial complications. Our aim was to reduce the length of hospitalization and assess safety and cost savings of discharging infants at a weight of 1900 g instead of 2000 g, as has been the practice.
Methods:
This is a single-centre, nurse led quality improvement project done at a tertiary neonatal unit in Singapore with primary outcome of reducing average length of stay in selected low birth weight infants. In phase 1, infants with birth weight between 1000 and 1700 g were discharged at 1900 g, provided they met the discharge criteria. Interventions were introduced in phase 2 after interim analysis for the two most common causes for delayed discharge: poor bottling skills and waiting time for scheduled herniotomy.
Results:
In phase 1, the mean hospitalization stay was reduced by 5.5 days, with 21% of the babies discharged at 1900 g. The safety of the intervention was assessed by rehospitalization rates, and found to be negligible. Interventions introduced in phase 2 to address the two major causes of delayed discharge did not improve the outcome. The estimated cost savings for each subsidized patient after implementation of the interventions was S$340–1100 over the two phases.
Conclusion:
Though only 21% of eligible infants could be discharged early, the study helped us identify key areas of intervention to facilitate early discharge of preterm infants. These included improving babies’ sucking skills, planning for early surgery, and providing adequate parental training. Safety and cost savings appear to be promising as well.
Keywords
Introduction
In Singapore, 9.2% of all children are born preterm, before they reach the 37th week of gestation. Of these, about 2% weigh below 2000 g at birth. 1 At KK Women’s and Children’s Hospital, we see about 11,000–12,000 deliveries per annum, with approximately 200 infants weighing less than 1500 g, putting them in the very low birth weight category. The practice at the hospital has been to discharge the babies who reached the 35th week of gestation when they attain the minimum post-birth weight of 2000 g. As a result, a significant proportion of low birth weight babies remain warded, even though they are medically fit to be discharged.
However, discharging babies before they reach 2000 g could be beneficial for the infant and family, as this allows parent–child bonding to begin earlier, while reducing the cost for healthcare services. There is a growing body of evidence demonstrating the feasibility and safety of discharging babies early.2–7 In a randomized control trial at an African hospital, the post-discharge health of babies discharged at 1650 g was comparable to babies discharged at 1800 g. In both groups, rates of weight gain, mortality, outpatient visits and re-admission to hospital were shown to be comparable. 3 In yet another study, Schmidt and Levine showed that babies who weighed 1800 g could be discharged safely when discharge criteria were met, with no complications or instances of morbidity post-discharge. 8 Notably, the study by Martín Puerto et al. study showed there was a 29% reduction in hospital expenditure when babies were discharged even before they reached 2000 g. 6 There were no differences in morbidity, mortality, growth or psychomotor development between discharged babies weighing under 2000 g and those who weighed more than 2000 g. 6
Our study aimed to find out if discharging low birth weight babies at 1900 g would result in shorter hospitalization. In addition, we wanted to explore the feasibility, safety and potential cost savings of discharging low birth weight babies at 1900 g if they were medically fit. A nurse-led workgroup was formed to oversee the discharge of infants when they reached 1900 g.
Methods
This study was a quality improvement (QI) project carried out in two phases. Baseline data was collected from March 2009 to August 2009 (phase 0). The first phase was from September 2009 to June 2011, and second was from July 2011 to December 2014. We included babies with birth weight of 1000–1700 g; who reached 35 weeks corrected gestational age; were born between September 2009 and December 2014; and did not have any significant medical concerns. When the baby reached 1900 g it would be eligible for early discharge if it met the following discharge criteria: i) he/she could bottle well, ii) could maintain a body temperature between 36.5 and 37.5°C in an open cot, iii) showed steady weight gain in the three to five days before the discharge date, and iv) had been apnoea-free for at least five days after discontinuing caffeine intake.
The primary outcome measure was the length of the hospital stay and the percentage of eligible infants that could be discharged at 1900 g. Secondary outcomes included evaluation of the reasons for delayed discharge and safety of early discharge was assessed by the number of readmissions in the early discharge group. This figure was obtained by phone conversation with parents and tracing electronic records. The nurse clinician recorded all information in a register.
The eligible babies were identified early and attended to by a dedicated nurse. The nursing team was involved in the planning, co-ordination and handling of the discharge procedures. Parents were taught how to care for their babies post-discharge, including administering cardiopulmonary resuscitation to the baby. Their competency in caring for their babies was assessed by a nurse-in-charge. The babies discharged at 1900 g were assessed after a week by a nurse clinician. Post-discharge growth, weight gain, head circumference, morbidity and re-hospitalizations were recorded. The babies that exceeded 2000 g were given BCG and hepatitis B vaccines. The parents were given the nurse clinician’s contact so they could contact her when any concerns arose. The project was started after institutional review board approval in 2009 and completed in February 2012. Discharging babies at 1900 g was established as the standard of care in March 2012 till date. We analysed records up to 2014. The only active interventions implemented in phase 1 were assessing the readiness of the baby for discharge at 1900 g. In phase 2, infants who showed poor suck at 33–34 weeks old were referred to the speech and language therapist, in addition to the phase 1 interventions. In phase 2, referral to the surgeon for herniotomy was recommended earlier, when the baby weighed 1800 g. Statistical analysis was done using paired Student’s t test. Statistical significance is marked when the p-value <0.05.
Results
A total of 791 infants with a birth weight ranging between 1000g and 1700 g were identified for early discharge from September 2009 to December 2014. There were 72 cases in phase 0 before start of programme. The entire cohort comprised 302 babies in phase 1 and 489 babies in phase 2. The baseline demographic and neonatal data for the babies in the three phases is captured in Table 1. The mean gestational age and birth weight of the babies across the three groups were comparable. The mean weight and gestational age at discharge and duration of hospital stay for all the three groups are given in Table 2.
Baseline characteristics of infants in all three study phases.
Babies who went home at <1900 g were discharged against physician’s advice. p values were not significant.
Mean weight and gestational age at discharge and duration of hospital stay for all three groups (phases 0–2).
Data is given for all babies born with birth weight 1000–1700 g during stated period. p values given for hospital stay using paired t-test. For discharge weight and gestational age, p values were not significant.
In the first phase, 302 infants were identified between September 2009 and June 2011. Of these, 23 infants were excluded as they did not meet the gestation criterion of 35 weeks. The 59 babies (21%) born with a mean birth weight of 1488g at the gestational age of 33.3 weeks were discharged at an average corrected gestational age of 37 weeks (±1.5), and at a mean weight of 1924 g (±18). The mean length of their hospital stay was 35 days. Their average weight gain was 35 g per day post-discharge. Four babies (6.7%) were readmitted for poor feeding, and discharged between one to three days. There was no major morbidity or mortality.
Among the remaining 220 (79%) babies, reasons for delayed discharge included poor sucking skills (82/220, 37%), herniotomy (46/220, 20.9%), parental request (30/220, 12.3%) and social causes (4/220, 1.8%). Medical reasons accounted for delayed discharge for 11.8% of the babies (26/220). These reasons range from anaemia requiring transfusions, apnoea of prematurity, necrotizing enterocolitis, or oxygen dependency requiring home care, and others. Four of the babies were discharged at under 1900 g, against the physician’s advice. The reasons for delayed discharge were not clear in 28 cases (12%), as seen in Table 3. After a preliminary analysis of phase 1 in June 2011, we started referring babies at 33–34 weeks with poor sucking skills to the speech and language therapist. We also started to refer babies to the surgeon for hernia repair earlier than usual.
Causes of delayed discharge.
At own risk (AOR) Babies discharge at less than 1900g at request of parents.
In phase 2, 489 babies weighing between 1000 and 1700 g were identified for early discharge, between July 2011 and December 2014. Thirty-two babies did not meet the gestational age criterion and were excluded, leaving 457 babies who were suitable for analysis. Of these 457 babies, 99 (21.6%) were discharged early. Their mean birth weight was 1496 g; their gestational age was 32.9 weeks. They were discharged after an average stay of 28 days, at a mean weight of 1921 g and gestational age of 37 weeks. The average weight gain at the first post-discharge visit was roughly 35 g per day. There were 358 (78.4%) babies that could not be discharged early, and of these poor sucking and swallowing skills was a major problem in 169 babies (47%). There were 34 babies (9.4%) awaiting herniotomy, 34 (9%) remained hospitalized at parental request, 49 (13.7%) had ongoing medical problems and 10 (2.8%) stayed on for social reasons. Seven out of 358 babies (1.9%) weighing under 1900 g were discharged against the physician’s advice, while 0.5% (2/99) of the babies discharged early were re-hospitalized. There was no reported death.
Fifty-five babies exceeded the discharge criterion weight of 1900 g while waiting to be discharged, and hence were included in category Other (Table 3). With improved planning and parental training, re-hospitalization rates have decreased, from 6.7% in 2011 to 0.5% in 2014. Similarly, parental request for delaying discharge decreased marginally from 12% to 9%. After implementing the early discharge programme, there was an average reduction of 5.5 hospitalization days per patient in phase 1 (p<0.001) and 1.7 days (p<0.001) per patient in phase 2 compared with phase 0. The dollar value of 5.5 patient days is S$1100, and S$340 for 1.7 days (average special care nursery cost of S$200 per day).
Number of babies that went home early as planned.
Among the very low birth weight (VLBW) subsets in both the phases, there were more small for gestational age (SGA) babies in phase 2. There was also a higher incidence of chronic lung disease at 36 weeks in phase 2 versus phase 1 (7.6% vs. 3.7%, p<0.05). The number of multiples and incidence of sepsis and necrotizing enterocolitis were comparable amongst the VLBW population in both phases.
Discussion
KK Women’s and Children’s Hospital is the largest provider of healthcare services for VLBW infants. The hospitalization cost for these babies is relatively high. Each additional day’s stay in special care nursery costs about S$200 for a subsidized patient; five times more for a non-subsidized patient. Standard hospital practice has been to discharge babies at a weight of ⩾ 2000 g, excluding babies who weigh less but are otherwise medically fit to be discharged.
When the project started, many babies were not discharged according to the weight criterion. This delay in implementing the new discharge criteria was due to lack of awareness resulting from poor planning. After issuing reminders to the nursing and medical staff and improving the identification process of the early discharge babies, the babies were discharged earlier according to the new criteria.
Prematurity is known to be associated with delayed discharge as elaborated by Klinger et al., who demonstrated that the most common causes for delayed discharge were prematurity-related complications such as broncho-pulmonary dysplasia, intra-ventricular haemorrhage, sepsis and necrotizing enterocolitis. 9 Merritt TA has summarized possible interventions that could help in shortening length of hospitalization. 4 At KK Women’s and Children’s Hospital, poor suck in babies was the major cause of discharge delay, and this was secondary to premature births in most cases. With the interim analysis in June 2011, we decided to seek the help of the speech and language therapist, referring infants with poor suck to the therapist at 33–34 weeks old. We initiated early referral for herniotomy – when the baby weighed 1800 g – resulting in a drop in the percentage of delayed discharge cases due to surgical reasons from 20% to 9%. Referral to a speech and language therapist did not contribute positively for early discharge in babies with poor suck. Of the 169 (47%) with poor suck, 75 (44%) remained in hospital awaiting clearance for discharge by the speech therapist.
There was a reduction of 5.5 mean hospitalization days in phase 1 (p = 0.001); however, phase 1 to phase 2 was not successful (p=0.08). Contributory reasons could be attributed to waiting for clearance from the speech therapist and increased number of babies with chronic lung disease and SGA babies.
We hope to bring down total hospitalization days further by introducing bottling to the babies at 32 weeks, or sending babies home on nasogastric feeding while awaiting good bottling skills. Parental request for delaying discharge decreased marginally from 12% to 9%. We can reduce this further by encouraging parents to spend more of their time in hospital caring for their babies to optimize their readiness for discharge.
The limitation of this study includes inadequate data collection. The reasons for delayed discharge were grouped only under broad categories. The major morbidities given above were only for VLBW infants (1000–1500 g) but our study included babies with birth weight range 1000–1700 g.
That said, this study has helped us understand problems that result in delayed discharge for preterm babies. We have since introduced a nurse-led early feeding pathway for non VLBW infants, and a speech therapist-led feeding pathway for VLBW infants. With these pathways, a limited percentage of infants who fulfil discharge criteria could be suitable for discharge at 1800 g. Future studies should focus on methods to improve bottling skills as this is a major reason for the delay in discharge.
Conclusion
Most of the infants could not be discharged at 1900 g as planned. However, a select group of infants could be safely discharged early in the presence of age-appropriate skills with early discharge planning. The feasibility of introducing bottling to the infant at 32 weeks and discharging with nasogastric feeding needs to be evaluated further.
Footnotes
Acknowledgements
I thank Ms Sheryl Quek Xianyu for editing the manuscript.
Declaration of conflicting interests
The authors declare that there is no conflict of interest.
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.