S. Aziz  ( Sindh Institute of Urology and Transplantation, Dow Medical College, Karachi. )
M. H. Rao  ( Sindh Institute of Urology, Dow Medical College and Civil Hospital and PMRC Research Centre Dow Medical College, Karachi. )

Objective:To determine the deficiencies in the existing record keeping system (ERKS) in government teaching hospitals of Karachi.
Study design:Descriptive, qualitative study to determine the deficiencies in the ERKS in government teaching hospitals of Karachi.
Setting:The study was conducted in four government teaching hospitals of Karachi from 1.1.1998 to 1.11.1998. Methods: A total of 9725 pre-coded proformas were filled from 136 units of all hospitals including out patient departments (OPD), wards and emergency units.
Results:Majority of the records (52.2%) were retrievable within one hour. The mean ± S.D of patients seen per day at all hospitals were 122.49±92.25(57 OPDs), 30.04±16.24(75 wards) and 186.25±145.80(4 emergencies). In majority of units (57.4%), the stationary for record keeping system was either not provided or was not enough. Only 39.0% units have 75% or more required information on records. The important information (except age) i.e temperature, pulse rate, blood pressure, diagnosis, weight, history of patient, laboratory findings, previous record and discharge report’ and prescribed doses of medicine were lacking.
Conclusion:The ERKS in the government hospitals of Karachi is inadequate. it is therefore essential that uniform OPD, ward and emergency slips should be developed having all-important queries. So that data could be utilized forresearch purposes along with betterplanning and management of the hospital and procurement plan for the government (JPMA 52:163;2002).

Karachi, one of the major cities of Pakistan, has a population of 9.8 million^1,2 though estimated to be: 13 million and an annual growth rate of 9-10%. There are foUr major government teaching hospitals in Karachi ‘with a substantial daily turn over of patients. (4600-690O)^3,4.
Published data from Pakistan on the availability of records from OPD, wards and emergency units are lacking or not of good quality. It is assumed that data from these
sources is not properly analyzed and not easily accessible in government teaching hospitals.
This study was done to assess whether the record keeping system in the government teaching hospitals is deficient and lacks quality or it is an assumption.

A descriptive qualitative study including interview, evaluation and observation was done at all departments of. government teaching hospitals of Karachi i.e., Civil Hospital (CH), Lyari General Hospital (LGH), Jinnah Postgraduate Medical Center (JPMC) and National Institute of Child Health (NICH) having 136 units including out patient department, wards and emergency.
The data was collected through three major pre­coded questionnaires i.e., Observation, Interview and Evaluation.
The interview questionnaire was filled from five persons where available at each unit. They were The whole data was collected by the investigators in each hospital by rotation, so that one investigator covered the wards of one hospital and OPD of the other hospital.
Ethical Consideration
The work was started after getting written permission from the Medical Superintendents of the respective hospitals who took permission from the Director General of Health Services if necessary. Medical Superintendents informed each head of the department through letter/circular. The investigators also took verbal consent from each Head of the units after providing verbal information concerning the project.
Statistical Analysis
Data recorded in the pre-coded proformas was stored in computer by means of EPI Info and SPSS 7.0 for Windows 98.

This study covered 57 OPD’s, 75 wards and 4 emergency units of the four government teaching hospitals of Karachi. Out of these, 55 were from CH, 46 from JPMC, 24 from LGH and ii from NICH. (Table 1).

From all above 136 units, a total of 9725 proformas (136 observation, 316 interview and 9273 evaluation) were filled for assessment of patients record, load of work, atmosphere and availability of other essential items.


The condition of the records kept in store rooms were also observed. Majority (80.9%) of records were safe and easily readable (Table 3).

The information on records was assessed according to the criteria laid down in the evaluation questionnaire. Only in 39.0% units, more than 75% information was available (Table 4).


The mean with S.D. of patients seen per day at each OPD was 122.49±92.25 in ward 30.04±16.24 and in emergency was 186.25±145.80 (Table 5).

The interview questions were put to different categories of staff. The information regarding stationary, place of storage availability and time required for retrieval of records, was obtained from all 136 units. The questionnaire was filled by 136 doctors, 81 nurses, 60 registration clerks, 30 ward masters and 9 record keepers. In most of the units (57.4%) stationary was either not available at all or was not enough (Table 6).

The working atmosphere was comfortable in most (58.8%) of the units (Table 7).


Majority of records (55.9%) were kept either in the store room or nurses duty room. The rest of the records were stored either at Deputy Medical Superintendent’s (DMS) office or statistical unit or in the storeroom of other departments.
As far as record retrieval was concerned, in majority units (61.2%), the record was available for one year or more. In 7.2% units, more than one month but less than one year record was available. In rest of the units, the record was either not available or shifted elsewhere (Table 8).


In all the hospitals, the speed for retrieval of records was good. In majority of units (58.8%) records could be obtained within a day (Table 9).

For evaluation purposes, 5% records from each OPD, ward and emergency were randomly selected.
In OPD’s of all the four hospitals (57 0 PD’s), a total of 7590 records were randomly selected. Out of these, 3231 were from CH, 2698 from JPMC, 1265 from LGH and 396
Majority of records (55.9%) were kept either in the store room or nurses duty room. The rest of the records were stored either at Deputy Medical Superintendent’s (DMS) office or statistical unit or in the storeroom of other departments.
As far as record retrieval was concerned, in majority units (61.2%), the record was available for one year or more. In 7.2% units, more than one month but less than one year record was available. In rest of the units, the record was either not available or shifted elsewhere (Table 8).
In all the hospitals, the speed for retrieval of records was good. In majority of units (58.8%) records could be obtained within a day (Table 9). For evaluation purposes, 5% records from each OPD, ward and emergency were randomly selected.
In OPD’s of all the four hospitals (57 0 PD’s), a total of 7590 records were randomly selected. Out of these, 3231 were from CH, 2698 from JPMC, 1265 from LGH and 396 Majority of records (55.9%) were kept either in the store room or nurses duty room. The rest of the records were stored either at Deputy Medical Superintendent’s (DMS) office or statistical unit or in the storeroom of other departments.
As far as record retrieval was concerned, in majority units (61.2%), the record was available for one year or more. In 7.2% units, more than one month but less than one year record was available. In rest of the units, the record
was either not available or shifted elsewhere (Table 8).
In all the hospitals, the speed for retrieval of records was good. In majority of units (58.8%) records could be obtained within a day (Table 9). For evaluation purposes, 5% records from each OPD, ward and emergency were randomly selected.
In OPD’s of all the four hospitals (57 0 PD’s), a total of 7590 records were randomly selected. Out of these, 3231 were from CH, 2698 from JPMC, 1265 from LGH and 396
Majority of records (55.9%) were kept either in the store room or nurses duty room. The rest of the records were stored either at Deputy Medical Superintendent’s (DMS) office or statistical unit or in the storeroom of other departments.
As far as record retrieval was concerned, in majority units (61.2%), the record was available for one year or more. In 7.2% units, more than one month but less than one year record was available. In rest of the units, the record was either not available or shifted elsewhere (Table 8).
In all the hospitals, the speed for retrieval of records was good. In majority of units (58.8%) records could be obtained within a day (Table 9). For evaluation purposes, 5% records from each OPD, ward and emergency were randomly selected.
In OPD’s of all the four hospitals (57 0 PD’s), a total of 7590 records were randomly selected. Out of these, 3231 were from CH, 2698 from JPMC, 1265 from LGH and 396
Majority of records (55.9%) were kept either in the store room or nurses duty room. The rest of the records were stored either at Deputy Medical Superintendent’s (DMS) office or statistical unit or in the storeroom of other departments.
As far as record retrieval was concerned, in majority units (61.2%), the record was available for one year or more. In 7.2% units, more than one month but less than one year record was available. In rest of the units, the record was either not available or shifted elsewhere (Table 8).
In all the hospitals, the speed for retrieval of records was good. In majority of units (58.8%) records could be obtained within a day (Table 9). For evaluation purposes, 5% records from each OPD, ward and emergency were randomly selected.
In OPD’s of all the four hospitals (57 0 PD’s), a total of 7590 records were randomly selected. Out of these, 3231 were from CH, 2698 from JPMC, 1265 from LGH and 396
from NECH. The mean with S.D. per OPD was 133.16±117.69. Age of the patient was available on 87.9% of the records. Age was recorded for 100% patients at CH and NICH, whereas at JPMC and LGH it was 87.7% and 53.6% respectively. The information regarding height and blood pressure was totally missing in all the hospitals. Specific or provisional diagnosis, which is a very important factor, was also missing in the majority (78.4%) of the records. At NICI I it was totally not recorded. The other important information i.e. weight, temperature, pulse rate, prescribed dose of medicine, record of previous visit and signature and designation ol’the doctor were also lacking at all the hospitals (Table 10).


At wards, a total of 417 records selected from all the four hospitals. (CH were randoml 198, JPMC 98,, ward was 5.56±5.14. Age was recorded for 100% patients at all the hospitals. All the other relevant information were available in majority of the cases (67.6% to 98.8%) except height and discharge report of the patients which was not available in most of the records (Table 11).

In the emergency units, a total of 1266 records were randomly selected from alt the hospitals. (CH 631, JPMC 408, LGH 109 and NICH 118). The mean ±S.D. per emergency was 3 16.50±251.49. Age was present on 91.4% of the records. Specific diagnosis was available only in NICH documents (100%). At other hospitals it was totally absent. No other information was available on any record of any hospital at emergency (Table 12),

At all the hospitals, no proper record keeping system was observed. The poor status of record keeping can be assessed from the fact that 136 units of all the hospitals, only 9 units have the post of a record keeper. Of these 8 were at JPMC and I at CH. This shows a lack of interest of the administration to develop a proper record keeping system. It was also observed in this study that more than 90% of the departments did not have a computer for record keeping and records were being maintained by pencil/pen on registers manually.
Only 39% units have 75% or more relevant information on records, due to which it could not be utilized for any planning or research purposes. Improved record keeping can only be done by the personal interest of the head of the units and management of the hospitals. As there is no such study available in the region, a comparison could not be made with this study.
it is concluded that basic information essential for records is missing at all the hospitals especially at OPD’s. No proper system was developed to keep records at all the hospitals. No proper place was available for record rooms.
Stationary was not available in sufficient quantity and good quality for maintenance of records. No trained person was appointed as record keeper in all the hospitals. Doctors do not give adequate attention towards documenting important information.

The authors are thankful to Pakistan Medical Research Council and Ministry of Health, Government of Pakistan for approval and funding this project under the Health Services Research Projects. The authors also acknowledge the help and cooperation of the Principals Prof. Dr. M. Shafi Qureshi, Professor Dr. Illahi Bux M. Soomro, all the Medical Superintendents/Directors and staff of the hospitals.

1.Economic Survey of Pakistan 1999-2000, Finance Division, Government of Pakistan.
2.Khattak,F.H. Health Economics and Planning in Pakistan, Ad.Rays Publishers, Islamabad,l 996, p1 0-28.
3.Aziz.S. Morbidity and Mortality at the neonatal care in a tertiary care hospital of Karachi, PJMR 1999;38:28-31.
4.Aziz S. Billoo AG, Samad NJ. Impact of socioeconomic conditions on Perinatal Mortality in Karachi. J. Pak. Med. Assoc., 2001;51: 354-360.