Table of Contents

HK J Paediatr (New Series)
Vol 1. No. 1, 1996

HK J Paediatr (New Series) 1996;1:77-81

Occasional Survey

A Survey on Pediatric Intensive Care Units in China

XM Fan, ZY Lu


To understand the present status of pediatric intensive care in China, we conducted a survey between October and December, 1993, involving 20 hospitals in 14 provinces and municipalities. The results showed that there were totally 41 ICUs for pediatric patients, including 19 pediatric ICUs (PICU), 18 neonatal ICUs (NICU) and 4 pediatric surgical ICUs (SICU), with total of 403 beds. The physicians to bed and nurses to bed ratios were 1:1.5 and 1:0.91 respectively. The average number of equipment per bed was 0.47 ventilator, 0.34 multi-function monitor and 0.47 infusion pump. Very few of the ICUs had portable X-ray machines, biochemical and blood gas analyzers, and hemodialysis machines. The most frequently treated diseases/conditions were pneumonia, intracranial infections, post-operation and sepsis in PICUs, and neonatal pneumonia, hypoxic ischemic encephalopathy and sclerema in NICUs. Pneumonia and respiratory failure accounted for 33.29% and 26.50% of all the diseases/conditions treated in the ICUs and the mean case fatality rate of respiratory failure was 24.50%. The results of the survey suggest that there is shortage of ICU beds and modern equipment, and treatment is often delayed due to excessively strict criteria for mechanical ventilation. A set of simple, and nationally acceptable criteria for evaluation of severity and cure of the diseases/conditions is urgently required.

Keyword : China; Intensive care units


Pediatric intensive care units (ICU) have been set up in hospitals in many provinces and cities since early 1980s, when pediatricians started to realize the importance of pediatric intensive care. Pediatric intensive care started in China relatively late, and many members of the medical profession are not familiar with its nature, model and objectives. To clarify the recent status of pediatric ICUs in China, we conducted a survey on 20 hospitals in 14 provinces and cities between October and December, 1993.

Materials and Methods


Forty-one ICUs in 20 hospitals established before 1992, which were endorsed by the Pediatric Group of Chinese Society of Emergency Medicine, were subjects of the present survey. These included 15 children's hospitals (which had connections with the Society) and 5 general hospitals in 16 cities listed below. All the hospitals involved in this survey located in large capital cities of provinces or municipalities except for Tangshan. There were totally 38 children's hospitals in China when the survey started, i.e., the 15 children's hospitals involved accounted for 39.5% of all the children's hospitals in China.


Nanchang Tianjin


Nanjing Wuhan


Shanghai Xining


Shenyang Xüzhou






Inquiry forms were filled in by the personnel in charge of the ICUs. The forms included date of establishment, size, number of beds, personnel structure, equipment, techniques applied and data regarding medical care of critically ill patients.


The time of establishment and structure of ICUs

Of the 20 hospitals, 8 established ICUs between 1982 and 1983; 5 established ICUs between 1984 and 1987 and 7 established ICUs between 1988 and 1992. There were 19 pediatric ICUs (PICU), 18 neonatal (NICU) and 4 surgical ICUs (SICU). Four hospitals had all three types of ICU; 13 hospitals had both PICU and NICU; 2 hospitals had only PICU and 1 hospital had NICU. In 5 hospitals the ICUs were incorporated in the department of emergency medicine; in 10 hospitals, the ICUs were independent departments, while in the others the ICUs were affliated to the departments of pediatrics, neonatology or surgery.

Number of beds in ICUs and the ratio of personnel to beds

In the 20 hospitals, the total number of ICU beds was 403, which accounted for 6.21% of the total number of pediatric beds (6,495) (5.44% in children's hospitals and 17.35% in general hospitals). The number of pediatric, neonatal and surgical ICU beds was 202, 176 and 25 respectively, giving a ratio of 1 : 0.87 : 0.12. There were 910 full time personnel, including 268 physicians; the physicians to bed ratio was 0.67 : 1. Among the physicians, the ratio of those with titles corresponding to professor (or associate professor), visiting doctor and resident doctor was 1 : 1.30 : 1.60. There were 442 nurses; the nurse to bed ratio was 1.1 : 1. Only a small number (0.81%) of nurses had medium title (similar to assistant professor) or above. Only a few of the ICUs had technicians or respiratory therapists.


Each ICU bed was equipped with an average of 0.47 ventilator (84.5% of which were imported) and 0.34 multi-functional monitor (of which 87.7% were imported). Among the NICUs, the average number of incubator per bed was 1.1 (14.7% were imported). In addition, each ICU had in average 1.1 transcutaneous PO2) monitors, 2.2 pulse oximeters, 2.2 blood pressure monitors and 1 electrocardiograph (ECG). Portable X-ray, automated biochemical analyser and hemodialysis were available in only very few ICUs.

Statistics of intensive care

  1. The sources of patients are illustrated in Table I. Most of the patients were admitted through outpatient department (OPD).

  2. Spectrum of diseases treated in the ICUs.
    The most common 10 diseases, which accounted for 60.3% of all admissions into the PICUs, are listed in Table II. Pnemonia (33.3%) and intracranial infections (7.60%) were the most common 2 diseases. Based on the data from NICUs in 10 hospitals, the most common 10 diseases were neonatal pneumonia, ischemic hypoxic encephalopathy, neonatal sclerema, intracranial hemorrhage, asphyxia, adult type respiratory distress syndrome (ARDS). hyperbilirubinemia, septicemia, low birth weight and congenital heart disease. A significant proportion of NICU patients were complicated by respiratory failure (26.5%), heart failure (15.4%) and intracranial hypertension (10.0%) Some infants also had hepatic (0.1%) and renal failure (0.4%).

  3. The therapeutic effects. The therapeutic effects of 56,676 infants and children treated in the ICUs are shown in Table III through VI and Figure 1. The overall mortality rate was quite high (Table III); however, the outcome of the patients seemed to have improved a little in 1992 as compared with that of 1984. This trend was also further demonstrated by reduction of mortality rates of patients who had undergone cardiopulmonary resuscitation (Table IV), mechanical ventilation (Table V), and of patients with RDS (Table VI). As shown in Table III through VI, in certain proportions of patients the ICU treatment was discontinued up on request of the parents. Those who died after abandoning treatment was not included in mortality. On the other hand, some parents requested for discharge when the patients' conditions had considerably improved but further ICU care was still needed. Among the low birth weight infants, the highest mortality rate was seen in those with birth weight lower than 1,000 grams (Figure).

Table I Sources of Patients
Sources Percentage
Emergency Room 46.8
Outpatient Department 34.6
Pediatric/Surgical Wards 15.8
Outside The Hospital 2.8


Table II The Most Common 10 Diseases Treated in PICUs
Diseases %
Pneumonia 33.3
Intracranial Infections 7.6
Postoperation 6.0
Septicemia 3.3
Poisoning 2.6
Congenital Heart Disease 2.3
Polyneuroradiculitis 2.0
Intracranial Hemorrhage 1.8
Airway Obstruction 0.8
Enteritis 0.6
Patients with these Diseases Account for 60.3% of all those treated in PICUs


Table III The Overall Therapeutic Effect
(%) (n=56,676)
  Total Mean Mean of 1984 Mean of 1992
Cure Rate 70.30 69.80 71.00
Improvement Rate 9.65 8.70 10.90
Treatment Given Up* 5.75 5.50 5.60
Mortality 14.30 16.00 12.50
Note. *Requested by Patients.


Problems in Organization, Structure and Management of ICU in Pediatrics

The availability and quality of intensive care can reflect the standard of medical care for critically ill patients in a country or a region. Although the number of ICU in China has gradually increased, the ratios of ICU to population and land area are still far less than those in European and North American countries. As an example, there are only 5 ICUs in Beijing with an ICU to population ratio of 1: 2.16 million and ICU to land area ratio of 1: 3361.8 km2, whereas in Paris there were 11 ICUs and the corresponding ratios were 1: 0.91 million and 1: 1090.9 km2, respectively. Therefore, pediatric ICUs should be supported for further development in China. In order to optimize organization and utilization of funds, the highest priority should be given to the children's hospitals at provincial and municipality level and those affiliated to medical universities/colleges (there are totally more than 40 such hospitals in China). Some pediatric intensive care beds should also be made available in some hospitals in large cities and even in some county hospitals. Integrated establishment of PICU, NICU and SICU will promote specialization of the ICUs, hence better care.

Pediatric ICUs in many other countries are independent units. In China, up to 50% of pediatric ICUs are affiliated to other departments, and this arrangement is affecting patient care due to difficulties in transfering patients and utilization of manpower and material resources. For example, the results of cardiopulmonary resuscitation of 162 cases in Beijing Children's Hospital showed that the success rate in ICU (53.5%, 46/86) was significantly higher than that in the wards of other departments (25.1%, 7/27) and the emergency room (36.7%, 18/49).1 Therefore, there should be guidelines concerning consultation and transfer of patients between ICU and the other departments. In 5 hospitals (25%) involved in this survey, departments of emergency medicine were set up by combining ICU and emergency room to form an emergency ICU (EICU). This might facilitate the care of critically ill patients in emergency room and OPD in a coordinated manner with ICU facilities. Hence EICU seems to be a novel model of intensive care.

Table IV Therapeutic Effect (%) of Cardiopulmonary
Resuscitation (N=2,345)
  Total Mean Mean of 1984 Mean of 1992
Cure Rate 38.2 28.4 39.2
Improvement Rate 9.3 NA NA
Treatment Given Up* 11.3 NA NA
Mortality 41.1 60.9 45.8
Note. *Requested by Patients. NA=Not Available


Table V Therapeutic Effect (%) of Mechanical
Ventilation (n=8,105)
  Total Mean Mean of 1984 Mean of 1992
Cure Rate 48.2 32.7 55.2
Improvement Rate 5.2 NA NA
Treatment Given Up* 13.7 NA NA
Mortality 32.9 49.4 31.3
Note. *Requested by Patients. NA=Not Available


Table VI Therapeutic Effect (%) on RDS
  IRDS (N=364) ARDS (N=555)
  Total Mean Mean 1984 Mean 1992 Total Mean
Cure Rate 38.50 12.50 48.00 60.36
Improvement Rate 10.20 NA NA NA
Treatment Given Up* 18.70 NA NA NA
Mortality 34.60 75.00 32.00 25.00
Note. *Requested by Patients. NA=Not Available


Fig. Thearpeutic effect (%) on low birth weight infants.

The optimal number of ICU beds should be 3% to 6% of total number of beds in a hospital. The present survey showed that the number of ICU beds was in accordance with this requirement. However, the ratios of physicians and nurses to the number of beds were obviously inadequate. In other countries the physician to bed ratio is usually 1-2 : 1, and the nurse to bed ratio 2.5-4 : 1. It is clear that in China, ICUs are staffed by fewer physicians and much fewer nurses. In most of the pediatric ICUs there are no special technicians or respiratory therapists; the large amount of specialized work normally performed by best trained personnel is inappropreately and unreasonably undertaken by physicians, a phenomenon that requires urgent rectification.

The results of the survey also showed that most of the pediatric ICUs are short of modern equipment, especially high quality ventilators. In many ICUs urgent blood gas analysis, blood electrolyte analysis and many other bed-side investigations are not available. The shortage of equipment is adversely affecting patient care and pose great disadvantages to research work.

Problems in Pediatric Intensive Care

Pediatric ICUs were set up in European and north American countries in 1960s for the management of mainly low birth weight infants and postoperative patients, especially those who required long term mechanical ventilation. The use of ICU has now been extended to cover critical diseases of various etiologies, and is playing an important role in reducing the mortality of critically ill infants and children. For example, in Saint Vincent de pau Hospital, Paris, France, the overall mortality rates in the pediatric ICU and NICU were 34% and 54% respectively in 1960s when these ICUs were just established, which declined to 15% and 18%, respectively, 10 years later.2 The present survey indicates that the introduction of pediatric ICUs in China has improved the total cure rate of critically ill infants and children to 70% and reduced the overall mortality rate to 14.3%. These figures however were comparable only to those in Europe and North America in 1970s. Furthermore, as the spectrum of diseases treated in the ICUs in China is obviously different from that in the European and North American countries, the cure rate and the mortality mentioned above may not reflect the real gap between the standard of pediatric intensive care in China and in developed countries.

The present survey indicated that the leading diseases treated in the pediatric ICUs were pnemonia and respiratory failure, which accounted for 33.29% and 26.5%, respectively, of the total admission. The mean case fatality rate of respiratory failure was 24.5% although it yarned among ICUs (range from 6.1% to 67.6%). The reason for such a high case fatality rate was that in some of the ICUs the criteria for tracheal intubation and mechanical ventilation were too strict and many patients were started on MV too late. It is critically important for clinicians to do away with the conservative belief that ventilator should only be used in dying patients and to pupularize and improve the techniques of MV.

MV is the most important therapeutic measure in ICU. The proportion of patients who received MV increased form 32.7% in 1984 to 55.2% in 1992 and the failure rate of MV declined from 49.4% (1984) to 31.3% (1992). The indications for MV were broadened from cardiopulmonary resuscitation to include respiratory, cardiac, circulatory, and cerebro-functional failures. Timely institution of respiratory support before respiratory and cardiac arrest had resulted in improvement in cure rate of IRDS from 12.5% to 48.0% and decrease in case fatality rate from 75.0% to 32.0%. The cure rate and case fatality rate of ARDS were 60.36% and 25.0% respectively. Treatment of IRDS with artificial surfactant in Pediatric Hospital of Shanghai Medical University, Children's Hospital of Hangzhou, and Capital Institute of Pediatrics, and the use of nitric oxide (NO) in treating ARDS in Beijing Children's Hospital had improved the therapeutic effects of MV. These therapeutic modalities however were not widely used in the country.

Although the number of patients who needed CPR accounted only for 4.1% of all patients treated in the ICUs, the success rate of CPR is one of the important criteria for evaluation of ICU care. The present survey showed that the success rate of CPR improved from 28.4% in the earlier years to 39.2% while the failure rate dropped form 60.9% to 45.8%. More aggressive application of tracheal intubation and better airway care are important differences in CPR between ICU and non-ICU departments. Large dose intravenous adrenaline (0.1 mg/ kg each time),3 intratracheal application of drugs (adrenaline, for example) for resuscitation, continuous infusion of vasoactive drugs (adrenaline, dopamine, aramin) through infusion pumps, aggressive ventilation treatment of cerebral edema etc. all helped improve the success rate of resuscitation.

In spite of the diversity of opinions in China at present on the value of treatment of low birth weight infants, the survival rate of these infants indeed reflects the standard of care of severely ill infants. Establishment of NICUs undoubtedly plays an important role. The present survey indicated that the survival rates of newborn infants weighing less than 1,000 grams, 1,001-1,500 grams, and 1,01-2,500 grams were 50.0%, 53.0% and 73.3%, respectively. These figures were comparable to those achieved by developed countries in the early 1980s. Improvement in NICU care with an aim to further reduce the mortality of low birth weight infants is necessary.

The Problems in Pediatric Intensive Care in China that Need Urgent Solution

  1. A set of objective, simple, practical, and comprehensive scoring system for critically ill children in China is needed for the evaluation of the severity of the patients' condition and for determining whether a patient requires ICU care. Such a scoring system is not only useful for the effective utilization of limited resources, but also can provide unified criteria for selection of cases and evaluation of therapeutic interventions in multi-center collaborative studies. Furthermore, it will be useful for quality control and scientific exchanges. The present survey showed great variation in patient outcome among the ICUs involved; e.g., the success rates of CPR in 1991 ranged from 34.0% to 92.1%; the case fatality rates of multiple organ failure varied from 10.7% to 62.7%. These discrepancies were probably due to the lack of unified criteria.

  2. Further discussions on "The Definition for Cure" are needed. Currently the criteria for ICU admission are based on organ functional failure, but cure rates are based on the nature of the primary diseases. Owing to financial constraints, shortage of beds and other reasons, patients are often transferred to other departments or discharged once their organ functions have recovered but before their primary diseases were cured. These patients could be listed among the cured and, as a result, the statistics cannot truly reflect the therapeutic effects or achievements of the ICUs.

  3. The present survey showed an increasing tendency for parents to abandon treatment of their children. Discontinuation of treatment was requested by the parents in over 5% of the total admissions, and for some critical diseases, the proportion was as high as 15%. The major reasons for giving up treatment include financial difficulty, female infants and poor prognosis. A scoring system for critical diseases may help estimate prognosis and provide relatively objective criteria for continuation or abandoning of treatment, and enable doctors to play a more active role in making these important decisions.

  4. Rational Use of ICUs. ICU is a necessary component of the clinical medical care. However, there are phenomena showing inadequate understanding of indications for transferring patients to ICUs. For example, some OPD doctors transferred less severe patients to ICUs; some ICU doctors also accepted such patients to avoid vacancy of their beds. On the other hand, some critically ill inpatients were transferred from different wards to ICUs at very late stage of their diseases, compromising their chance of rescue. In addition to setting up and application of a unified scoring system for critically ill patients as mentioned above, proper education of physicians working in different department in terms of rational use of ICUs is also required for optimal utilization of the equipment and techniques of ICUs.


1. Li K. Analysis of the results of cardiopulmonary resuscitation of 162 cases (with follow up of 41 cases). J Practical Pediatr 1993;8:55-6.

2. Fan X. Introduction to the pediatric intensive care units and the transfer center of critically ill children in Paris, France. Chinese J Pediatr 1983;21:182-3.

3. Zhu Y The effects of large dose adrenaline on cardiopulmonary resuscitation in children. J Practical Pediatr 1993;8:264-6.


This web site is sponsored by Johnson & Johnson (HK) Ltd.
©2024 Hong Kong Journal of Paediatrics. All rights reserved. Developed and maintained by Medcom Ltd.