STANDARDS FOR NURSING DOCUMENTATION IN GENERAL HOSPITALS IN SOUTH AFRICA

INTRODUCTION AND PROBLEM STATEMENT The setting of practice standards for South African nursing has emerged as a research priority since 1985 (SANA). The setting of standards is the first step in quality assurance programmes. The steps of a quality assurance programme usually consists of the following: 1. setting of standards based on values; 2. formulating criteria to measure each f standard; developing measuring instruments; measuring the quality of care; 5. identifying problem areas; 6. designing and implementing remedial action; 7. remeasuring the quality of care (Bruwer, 1987). In the regulations of the South African Nursing Council it is clearly stated that recording the course of the patient’s health problem, the care received and the results of this care, is the responsibility of the registered nurse (Reg. nr. 2598 of 30 Nov 1984). When a project was started in 1987 to develop standards for nursing in general hospitals, it was decided that the recording aspect of practice should be a priority for the following reasons: 1. Most of the available quality assurance instruments are based at least partly on chart reviews. Badly kept records thus makes any quality assurance «ogramme difficult. Nurses in South Africa have been endeavouring to implement the problem oriented approach to their records, and this has focussed a lot of attention on record-keeping. 3. Some nursing research has shown that this aspect of nursing practice is not of a satisfactory quality (Booyens, 1987). The objectives of this study were: 1. To formulate standards for all nursing records, valid for all medical and surgical units in all types of general hospitals in South Africa. 2. To design an evaluation instrument based on these standards; 3. To identify factors which influence nursing docum entation significantly, and draw up expectancy tables with regard to these factors in order to make comparisons possible.

f standard; developing measuring instruments; measuring the quality of care; 5. identifying problem areas; 6. designing and implementing remedial action; 7. remeasuring the quality of care (Bruwer, 1987).
In the regulations of the South African Nursing Council it is clearly stated that recording the course of the patient's health problem , the care received and the results of this care, is the responsibility of the registered nurse (Reg. nr. 2598of 30 Nov 1984).When a project was started in 1987 to develop standards for nursing in general hospitals, it was decided that the recording aspect of practice should be a priority for the following reasons: 1. Most of the available quality assurance instrum ents are based at least partly on chart reviews.Badly kept records thus makes any quality assurance « ogramme difficult.
Nurses in South Africa have been endeavouring to implement the problem oriented approach to their records, and this has focussed a lot of attention on record-keeping.
3. Some nursing research has shown that this aspect of nursing practice is not of a satisfactory quality (Booyens, 1987).
The objectives of this study were: 1.To form ulate standards for all nursing records, valid for all medical and surgical units in all types of general hospitals in South Africa.
2. To design an evaluation instrument based on these standards; 3. To identify factors which influence nursing docum entation significantly, and draw up expectancy tables with regard to these factors in order to make com parisons possible.the association between good patient care and good record keeping.Phaneuf (1976: 48-49) says in this regard that research has shown a positive correlation between the quality of medical records and the quality of medical care, and that this is also true about nursing, since "the conditions that bring about good care are also responsible for bringing about good recording".

LITERATURE SURVEY
Nursing records cover all three phases of the patient's hospitalisation, namely admission, progress and discharge.In a study by Ciuca (1972) on 235 nursing care plans in six hospitals, it was found that 72% of the recordings dealt with medication, treatm ent, m onitoring of vital signs, intake and output and diagnostic tests, while only 28% dealt with nursing interventions.
Most of the quality assurance instrum ents currently in use is based at least partially on auditing of records.Phaneuf's audit instrum ent is a retrospective inspection of a nursing record (1976).M onitor, a British instrum ent which was based on the American Rush-Medicus instrum ent, is a comprehensive list of 220 criteria, which is assessed either through a chart review or interviews or observation (Goldstein e.a. 1983).All these instrum ents are focused on process standards, as opposed to structure and outcome standards.

TERMINOLOGY:
• Nursing records or nursing docum entation is the inscriptions of registered (including students) or enrolled nurses in the patient record during admission, treatm ent and discharge.• A standard is a statem ent describing the expected level of perform ance against which the quality can be evaluated.• A criterium is an indicator which is suitable for measuring a standard.

METHODOLOGY:
The study was carried out in the following steps: • The form ulation of standards for docum entation.• The development of an evaluation instrument.• The identification of the significant factors influencing nursing docum entation.

The formulation of standards for documentation:
A thorough literature survey was done and three general standards were then form ulated, together with the criteria for each.
The three general standards were the following: a) The docum ent complies with legal requirements.b) The docum ent is a complete record of the condition of the patient and the nursing care rendered.c) The docum ent is an effective record of reality.The Nursing Record Standard Sheet was then presented to two groups of registered nurses from different regions in the country.They were required to evaluate the standards and criteria according to the following questions: • Is there consensus on the meaning and clarity of each item?• Is the criterium an acceptable indicator of the standard of care?• Should any criteria be deleted or added?• How im portant do they judge the criterium to be?
The evaluation first took place individually, and was then debated by the group.
This peer group evaluation lead to the wording of many criteria to be am ended, criteria being left out because consensus could not be reached on their acceptability, C urationis Vol. 12, Nos. 1 & 2, July /Ju lie 1989 and some criteria being added.On the basis of the work of the peer groups, the content validity of the Nursing Record S tandard Sheet was accepted.

The development of an evaluation instrument:
2.1 Instrum ent: The criteria which had been identified were arranged in the form of a check list which could be used to audit a patient record.In some cases the d ata in the record had to be validated against other docum ents, e.g.d octor's notes, flow charts or by questioning the patient.
The instrum ent was prepared so that the data could be com puterised.A small pilot testing of the instrum ent was done in one hospital.

Sam ple:
The instrum ent was tested on a stratified random sample in two regions of the country, as illustrated in Table 1.In each hospital the w ards/units to be used were random ly chosen.In every unit 20% of the records were random ly chosen to make up a total of 10% of the total num ber of beds in each hospital.Intensivecare units, casualty-, outpatient-and midwifery departm ents and operating theatres were excluded.
2.3 F ield workers: Three registered nurses were used as field workers, after receiving training from the researchers.

Independent variables:
It was expected that certain im portant variables might influence the quality of docum entation.It was therefore planned to collect inform ation about the following independent variables:

Validity:
• Content validity.This was accepted since the instrum ent was based on valid standards and criteria.• C onstruct validity.It can be assumed that the quality of nursing docum entation will vary with both the bed occupancy of a unit and the staffing levels.This construct was tested with the Nursing Record Standard Sheet (NRSS) and a significant difference (p = 0,000) was found between the quality of docum entation as measured by the N RSS and these two variables, using a Chi-square correlation analysis.This could be seen as support for the argum ent that the N RSS is a valid measure of the quality of nursing docum entation.

Reliability:
Inter-rater reliability was calculated on 30 records that were audited by two field workers.The Chi-square correlation could only be calculated for 40 items, and a correlation of between p = 0,7970 and p = 1,000 was found for 70% of these items.On only three items were a low correlation (p = 0,05) found.These items required the field workers to count things like the num ber of notations and the num ber of signatures over many pages.In the final instrum ent these items were changed to limit the counting, and so increase the reliability.

The identification of the significant factors influencing nursing documentation:
It was found that the following three variables made a significant difference (p = 0,000 by means of C hi-square analysis) to the quality of docum entation: • the percentage of registered nurses per unit • the percentage of non-nursing support staff per unit • the bed-occupancy rate per unit The effects of the percentage of registered nurses and the bed-occupancy are summarised in Tables 3 and 4 in the form of expectancy tables.These tables can be used by users of the N RSS to allow for these variables when interpreting results from their own settings.
The influence of the hospital in which the docum entation was done, was calculated through an ANOVA and was found to be highly significant (p = 0,000 with F = 29,922 with 458 degrees of freedom).The total quality of the docum entation in the different hospitals is given in Table 5.It is interesting to note that the two hospitals which had an average docum entation quality of less than 3,00 (hospitals 3 and 8), were the only private hospitals in the sample.The next lowest, hospital 1 with 3,02, is a large Black hospital with a much higher bedoccupancy rate and lower staff numbers than hospitals 3 and 8.
The type of unit also made a significant difference (p -0,000) to the quality of docum entation (ANOVA: F = 5,104 with 458 degrees of freedom), but the influence is difficult to interpret (see Table 6).The instrum ent, its cue sheet and instructions for adm inistering it is available at the publication section of the SA Nursing Association.
Although this instrum ent evaluates only one aspect of the practice of nurses in general hospitals, it should be a useful tool in the hands of nurse adm inistrators.
, A. 1981 Q u a lity c o n tro l in nursing.U n p ublished D. C u r d isse rta tio n , U P E .B ooyens, S.W .1987.D ie toepassing van die w e te n ska p like m e to d e in verpleging in sekere S u id -A frik a a n s e h o sp ita le -gehalte e valueringsperspektief.U n p ublish ed D .L itt. et Phil d isse rta tio n , U N IS A .C iuca, R .I.. 1987.O ver th e years w ith th e n u rsin g care plan.N u rsin g O u tlo o k , 20 (11): 706-711.G o ld sto n e , L .A ., Ball, J .A ., C ollier, M .M .1983.M o n ito r, an in d e x o f th e q u a lity o f n u rsin g care f o r acute m ed ica l a n d surgical wards.2nd ed.N ew castle U pon T yne, N ew castle U pon T yne P olytechnic.P a to n , F. 1986.P rio rity areas fo r n u rsin g research.C urationis 9 (2): 13-14.P h an eu f, M .C .1976.The n u rsin g a ud it: se lf regulation in n u rsin g practice.N ew Y ork: A p p leto n -C en tu ry -C ro ft.S earle, C ., R o b e rtso n , B. and Nel, C .M .1985.V erpleeg-adm inistrasie: S tu d ieg id s I vir N V A 1 0 0 -L .P re to ria : U N IS A .This research was su p p o r te d b y a H S R C grant.We w ant to th a n k th e fo llo w in g p e o p le f o r their assistance at d iffe re n t stages o f th e research: M rs L. C oetzee, D r A .Bruw er, D r C. W. F ólcher a n d M iss A .van Bruggen.
.N ., R .M ., Dip.In ten sive Care A sso c ia te P rofessor, D ep a rtm en t o f N ursing, U niversity o f S o u th A frica .
L .R .U ys D .S o c .S c ., R .N ., R .M ., P .N ., D .N .A ., D .H .A P rofessor a n d H ead, D ep a rtm en t o f Nursing, U niversity o f N a ta l S. W. B ooyens, D. L itt et P hil ( U N I S A ) R