Territory-wide Obstetrics and Gynaecology Audit (1994)

Background

Over the past few decades, hospitals providing obstetric and gynaecological services had been collecting data about deliveries and gynaecological conditions in varying format and complexity. But the data collected were different, and sometimes even the definitions might vary from hospital to hospital, making it difficult to produce an overall view of the obstetrical and gynaecological scene in Hong Kong. With the development and widespread use of computers, the data collected in various hospitals could be amalgated together and a more complete picture in Hong Kong presented.

Hence, in early 1992, the Hong Kong College of Obstetricians and Gynaecologists (HKCOG) had called the representatives from the training hospitals together, and a working group was formed to look into this issue. Data items being collected by the participating hospitals were examined item by item, so that essential information would be collected, while not overloading the doctors who had to collect the information. Finally, a 'Minimal Data Set' (MDS) with definitions of terms used were agreed upon. Every hospital would collect these data, but they could collect more than the MDS if they desired. A computer program was developed by Professor Allan Chang for data collection and report generation. After a trial run of` 2 months, the program was implemented in January 1993.

In 1993, out of the 9 training hospitals, 6 had used the obstetric program and 7 had used the gynaecological program, while the rest continued to use their pre-existing programs. In July 1993, St. Paul's Hospital conducted a trial run of the program, which was so successful that the Quality Assurance Committee of HKCOG decided to invite all the private hospitals (through the Data Collection Subcommittee and the co-ordinators) and the other public hospitals (through the Computerised Statistical Audit Subcommittee) to join in using the program in 1994.

The 1994 Territory-wide Audit

In 1994, 22 hospitals providing obstetric and/or gynaecological services in Hong Kong had participated in using the program. In the 11 public hospitals, the audit forms were filled in and data entry was completed in the hospitals, and the database in floppy disks were sent to the College. For the private hospitals, the filled audit forms were sent to the College for centralised data entry at regular intervals.

After the first three months' data were collected and pooled together, the hospitals and the patients' identity were eliminated and a preliminary report was compiled for discussion in the Clinical Audit Subcommittee, so that the format of the report would be both useful and easily interpreted. At the same time, the data entry for the rest of the year was being continued.

Data Processing

After pooling all the data obtained in the whole year of 1994, the hospital names were replaced by codes, so that the hospitals would not be identified. The patients' identity card numbers (if present) were trimmed to 5 characters (usually 1 character plus 4 numerics). Various fields were validated. Duplicated records were eliminated if known. A combined report of factual description was produced, and some topics of interest were analysed.

The vital statististics were checked against the data published in the Hong Kong 1995 - A review of 1994 (published by the Hong Kong Government) and discrepancies in the numbers of stillbirths and neonatal deaths were noted.

Since some of the babies were transferred to paediatricians for further care and outcome might be unknown at the time of filling the audit form, various paediatric departments were approached for further information before compiling the final report. A further 39 neonatal deaths were identified and added to the original data.

Obstetric Report

In 1994, the total number of known live births and stillbirths in Hong Kong was 71,989 (Hong Kong 1995 - A review of 1994) and the College had received 68,083 obstetric returns, making a return rate of 94.6%. As this was a voluntary participation, the 5.4% missing data might be significantly different from the rest of the audit, and was unknown. Some of the data items might have been cross-checked, but it was impossible in most of the fields and the accuracy depended on the doctor who filled the form initially.

A default value was set for the commonest condition for most items of the obstetric part, this facilitated data entry, but also made omission of data difficult to identify.

At the time of entering data into computer, there was an error-checking process to check the completeness and validity of certain data. The record would be marked as 'Perfect', 'Incomplete' or 'Erroneous' accordingly. Upon checking the obstetric data, 1.8% of records had 1 or more items missing, and 10.2% had inconsistent data in 1 or more items. One of the problem of centralised data entry was that when the inaccurate data were identified, it was long after the event and the data could not be traced, while the doctor who filled the form was also unaware of the error.

Out of the 202 neonatal deaths registered (Hong Kong 1995 - A review of 1994), 132 were included in the initial analysis, and after further tracing from various paediatric departments, 171 (85%) were included.

This report was based on the data collected in the survey and great effort was made to entertain the validity, but 100% accuracy could not be guaranteed. Some of the data were well-defined and should be fairly accurate (such as age, birth weights, mode of delivery etc), but some would need some interpretation (such as duration of labour, gestational age). However, the fetal outcomes, neonatal complications etc should be interpreted with care because it depended very much on the effort of the obstetricians to trace the outcomes and co-operation of the paediatric colleagues to give the feedback, hence the neonatal complications

Gynaecological Report

For the gynaecological part, only hospital admissions were included and the College received 60,809 returns. There was no known figures about gynaecological admissions in Hong Kong, the return rate could not be calculated. Upon checking, 9.2% of the records had 1 or more missing data.

Owing to the initial desire to keep confidentiality of patients, the patients' identity card number was an optional item, resulting in duplication of diagnoses since the data in the gynaecological part were admission-based. Hence, the true incidence of diseases could not be accurately known. The incidences quoted in this report should be interpreted with caution. In addition, since the default value of complications were set to 'Nil', these might be under-reported and should be interpreted with caution - but the figures should be the minimum incidence.

Conclusion

This report is the first of its kind in Hong Kong and great effort had been spent in its compilation. Valuable experience was gained and more useful information would be collected in the next exercise.

LISTS OF WORKING GROUPS AND SUBCOMMITTEES

Working Group on O&G Audit (1992-93)

Standing Committee on Quality Assurance (1993-96)

Dr LEUNG Pamela  (93-96, Chairman, 93-94)
Prof HO Pak Chung  (Chairman, 95-96)
Dr MULLINS Lewis  (Co-chairman, 93-94)
Dr CHANG Fuk To  (93-96, Co-chairman, 95-96)
Dr CHIN Robert  (Hon. secretary, 93-96)
Dr CHAN Joseph WT  (93-96)
Dr CHAN May YM  (93-96)
Prof CHANG Allan MZ  (93-96)
Dr CHEUNG Tak Hong  (95-96)
Dr HAINES Christopher (93-94)
Dr HO Lau Cheung  (93-96)
Dr SO William WK  (94-96)
Dr TAI  Chark Man (95-96)
Dr TANG Kwok Hung  (93-96)
Dr TANG Lawrence CH  (93-96)
Dr WONG Tung Choy  (1993)
Dr WONG-TAM Vivian CW  (93-96)

Data Collection Subcommittee (1993-95)

Dr CHANG Fuk To (Chairman)
Dr CHAN Joseph WT 
Dr CHAN May YM 
Dr FOK Nancy LL 
Dr TANG Kwok Hung 
Dr TANG Lawrence CH 
Dr WONG Tung Choy

Private Hospital Co-ordinators:
Dr Chan Joseph (Hong Kong Santorium & Hospital Ltd)
Dr Chang FT (St Paul 's Hospital)
Dr Fung KK (St. Teresa Hospital)
Dr Lam CC (Evangel Hospital)
Dr Lee Eric (The Hong Kong Central Hospital Ltd.)
Dr Lee KW (Precious Blood Hospital)
Dr Lo Peter (Tsuen Wan Adventist Hospital)
Dr Mullins L (The Matilda & War Memorial Hospital)
Dr Soo HN (Canossa Hospital)
Dr Albert Tsai (Hong Kong Adventist Hospital)
Dr Patrick Wong (Hong Kong Baptist Hospital)

Computerised Statistical Audit Subcommittee (1994-95)

Dr CHAN May YM (Chairman)
Dr TSE Kai Tai (Secretary)
Dr FUNG Hedy YM
Dr HO Pak Chung
Dr JONG Alexander LF
Dr LEUNG Belinda FH 
Dr MOK Ka Ming 
Dr NG Tai Keung
Dr PANG Chung Pui
Dr PAU Wing Fu
Dr PUN Ting Chung
Dr SIN Sai Yuen
Dr TSANG Sing Wing

Clinical Audit Subcommittee (1995-96)

Dr CHAN May YM (Chairman)
Dr TSANG Sing Wing (Secretary)
Dr CHAN Joseph WT
Dr CHUNG Ka Leung
Dr FOK Nancy LL
Dr FUNG Hedy YM
Dr LEUNG Pamela
Dr LI Dominic FH
Dr PUN Ting Chung