Response times of Ambulances to Calls from Midwife Obstetric Units of the Peninsula Maternal and Neonatal Service (pmns) in Cape Town Correspondence Address

Curationis 32 (1): 59-66 MOU, am bulance, response tim e, Response times of ambulances to calls from Midwife Obstetric Units, although var-^MNS jc(j, are perceived as slow. Delays in transporting women experiencing complications during or after their pregnancies to higher levels o f care may have negative conse­ quences such as fetal, neonatal or maternal morbidity or death. An exploratory descriptive study was undertaken to investigate the response times of ambulances of the Western Cape Emergency Medical Services to calls from mid­ wife obstetric units (MOUs) in the Peninsula Maternal and Neonatal Services (PMNS) in Cape Town. Response times were calculated from data collected in specific MOUs using a specifically developed instrument. Recorded data included time o f call placed requesting transfer, diagnosis or reason for transfer, priority of call and the time of arrival o f ambulance to the requesting facility. Mean, median and range o f response times, in minutes, to various MOUs and priorities of calls were calculated. These were then compared using the Kruskal-Wallis test. A comparison was then made between the recorded and analysed response times to national norms and recommendations for ambulance response times and maternal transfer response times respectively. A wide range o f response times was noted for the whole sample. Median response times across all priorities o f calls and to all MOUs in sample fell short o f national norms and recommendations. No statistical differences were noted between various priorities o f calls and MOUs.


Introduction
The Peninsula Maternal & Neonatal Service (PMNS) was established in 1980 by the Department o f Obstetrics and Gynaecology at the University o f Cape Town with the support of the then Cape P rovincial A d m in istratio n (Van Coeverden De Groot, 1993: 225).The service provides "tiered, comprehen sive" care to pregnant women and neonates within the Cape Peninsula.The PMNS consists o f primary care facilities known as Midwife Obstetric Units (M OU's), two secondary care facilities namely Mowbray Maternity Hospital (MMH) and New Somerset Hospital (NSH) while Groote Schuur Hospital (GSH) provides tertiary level care.28 000 births occur within the PMNS annually; more than 50% of the births in this service are managed at MOU's with 33% of births occurring in secondary facilities and 15% o f the births in the PMNS managed at a terti ary level o f care (Van Der Spuy, 2003).The service has protocols for clinical management and care which are shared by all the maternity care facilities within the PMNS, as well as a system o f refer ral between the various levels of care.Referral is facilitated by developed protocols, communication between the various facilities and the availability of transportation to transfer mothers and neonates.This study looks at the Figure 1: PMNS referral structure response times of ambulances in this setting.The National Committee for the Confidential Enquiries into Maternal Deaths (NCCEMD) recommends that emergency transport be available for all antenatal and peripartum women and their babies who experience complica tions.The indicator suggested is the time it takes for an ambulance to arrive on the site from where it was sum moned within one hour.This recom mendation has been the motivation for this study.

Background
Midwife Obstetric Units are "decentral ized obstetric delivery centres" devel oped by Philpott in the then Salisbury, Rhodesia now called Harare, Zimbabwe (Gunston & De Groot, 1975:1147).At this level o f care uncomplicated preg nancies and births are managed.At the core o f the functioning of the MOU is the registered midwife who performs the initial risk assessment and only keeps "low risk" patients for antenatal, intrapartum and postnatal care at the facility (Van Coeverden de Groot, 1993: 225).Midwives at MOUs are expected to be "competent in detecting abnor m a lities an d p ro v id in g em ergency c a re ", if necessary, before referring high-risk patients to the next level of care.The MOU also functions as a first "port of call" in an emergency (Van Coeverden de Groot, 1993: 225) if and when experienced by pregnant women in the community.MOUs have no resi dent medical staff (Van Coeverden de Groot, 1993:225) and midwives at these facilities depend on the Emergency Medical Services (EMS) to transport high risk patients and emergencies needing higher levels of care.
There are presently six MOU's in the PMNS providing antenatal, postnatal and 24 hour labour care to low risk mothers and neonates.Three o f these run satellite clinics for antenatal and postnatal care.Staff from the MOUs service these clinics for specific ses sions during the week.Since this study was conducted, one satellite facility has been closed due to staff shortages.All the facilities have specified referral cen tres to which they refer high-risk pa tients, complications and emergencies.Figure 1 shows the referral system structure.
The guidelines for the referral o f pa tients include a system of prioritizing calls for the transfer of women in the peripartum period.At the time that this study was conducted, the most recent document explaining the various pri orities of calls was departmental notice 27/1987 of the University of Cape Town Department of Obstetrics and Gynae cology.(Department of Health, 2002).The indicator to be used to assess the fulfillment of this recommendation is the time taken from the placement o f the call to the arrival o f the ambulance at the facility from where the patient is to be collected.The target to have been reached by December 2004 was 50% of ambulances to arrive at the facility within one hour o f the call being made (Department of Health, 2002).The mo tivation for this is that transport de lays were directly related to a signifi cant proportion o f maternal deaths.This report further asserts that trans port problems between facilities con tributed to 6.6% percent o f the assess able maternal deaths in the Western Cape during the triennium 1999-2001(Department of Health, 2002).This fig ure was one of the lowest, after the Free State, compared to other provinces for this particular avoidable factor.The Cape Peninsula is particularly well provisioned in terms o f infrastructure such as roads and communication as well as the existence o f the PMNS sys tem and protocols.

Responsibility for the provision o f emergency medical services through
out the Western Cape lies with the pro vincial government (Veriava & Cumow, 1997: 6).The Cape Peninsula forms the most south-western part o f the West ern Cape.Media reports have sug gested that delays are being experi enced in the response of ambulances to calls (Smetherham, 2004).This report alluded to the delays being a result of dire budgetary constraints being expe rienced by the EMS.It has been as serted that the restructuring in the health system has necessitated patients being transferred from one facility to another and that 40% o f ambulances spend th eir tim e undertak in g interhospital transfers (Veriava & Cumow, 1997:7).The assumption made here is that not all interhospital trans fers constitute emergencies requiring an ambulance.The use o f the emer gency medical services for the transfer o f patients being referred for tubal ligation also needs to be reconsidered.
The migration o f people seeking health care in the province has significant cost and administrative implications for the health department in the Western Cape (Human, Kroon, Bergman, Fawcus & Ntwana 2003: 634).This means that there are increased numbers of people, not necessarily permanently resident in the Western Cape, who rely on and therefore add to the workload of the EMS.Mongwe asserts that people are seeking help in Western Cape Health facilities, because of the perception that a "higher level of service" exists when com pared w ith the E astern Cape (2002).This adds to the load that needs to be carried by health care services in the Western Cape including the emer gency medical services.In addition violence is being directed towards EMS personnel in certain areas (Kassiem, 2005).This means that staff subjected to acts o f violence in the course o f their duties, may have to be removed from response duties in order to deal with these violent acts directed at them.This influences the number o f personnel available to staff vehicles adding to the workload o f an over-burdened service.
Anecdotal reports have shown a wide range of response times by ambulances to calls made from MOUs to transport patients as referrals to other facilities providing higher levels of care.Re sponse times have been acknowledged by Dr. Cleeve Robertson to be a "criti cal issue" (Smetherham, 2004).

Problem Statement
Response times of ambulances to calls from MOUs, although varied, are per ceived as slow.Delays in transporting women experiencing complications during their pregnancies, births or in the postnatal period to higher levels of care may have negative consequences such as fetal, neonatal or maternal mor bidity or death.

Aim of the study
The intention of the study was to de termine the response times of ambu lances to maternity and neonatal spe cific calls from MOUs within the PMNS.

1.
To establish the response times of ambulances to calls from MOUs in the PMNS for all (Obstetric and Neonatal) transfers.2.
To compare the ambulance response times to various MOUs.

3.
To establish the response times of the various priorities of calls i.e. ordinary ambulance, urgent ambulances and "flying squad".4.
To compare ambulance response times with the target set by the 'Saving Mothers' report of the National Depart ment o f Health.

Literature review
S earches using PubM ed® and Biblioline® were made to determine what previous work has been done on the response times o f ambulances to maternity care facilities and on ambu lance responses to domestic calls and the pre-hospital environm ent.The search terms used were: "PMNS", "am bulance response times", "maternal transfer" "Cape Town" , "obstetric emergency" and "peninsula maternal and neonatal service" Many o f these studies looked at the implications of quick responses on mortality and mor bidity in settings other than maternity care.The only published work that could be found relating to the re sponses of ambulances to interhospital transfers in the maternity setting, was by Van Coeverden De Groot, Van Coeverden De Groot, Smith and Isaacs in 1994 who reported that ambulance delay was "disturbingly common" in transferring patients from MOUs with poor progress in labour.They asserted that ambulance delays are the most sig nificant contributors to the delay expe rienced in the transfer o f patients from MOUs to hospitals.The findings re late to the length of time between "sum moning the ambulance to the MOU and the patient's first assessment at the hospital".These delays were compared to hospital procedural delays (Van Coeverden De Groot et al, 1994:73).The study did not assess the length o f time it took from the first call to request an ambulance to the arrival of the ambu lance at the facility.
Gibson, Bailey and Ferguson (2001:300) assert that transporting mothers with fetuses in-utero to higher levels of care can be regarded as a key strategy in improving the survival o f high-risk in fants.It has also been argued that the use of guidelines for the transfer/refer ral of pregnant women improves the outcome for babies in terms of morbid ity and m o rtality (G ibson e t a l, 2001:300).The use of protocols is a key feature of the PMNS.

Methodology
An exploratory descriptive study was undertaken to investigate the response times o f ambulances to MOUs within the PMNS.This study design "pro vides the least control over variables" (Brink & Wood, 1998:283) because "the situation, not the investigator, controls the data" (Brink & Wood, 1998:285)

Study sample
The

Instrument
Data was collected by means of a tabu lated log sheet.The instrument re quired the recording o f the time a call was placed to the ambulance control room to transfer a patient, the reason or diagnosis, the priority of the call in terms o f whether an ordinary, urgent or flying squad was requested and the time o f the arrival of the ambulance.Space was allocated to record the last three/four digits of the patient folder number.

Validity
The measurement o f the time it takes for ambulances to respond to calls placed from the MOUs on the instru ment ensured face validity.Clocks are valid instruments for measuring time.
Each MOU had available, working clock/s.The instrument therefore meas ured what it was supposed to measure (response times o f ambulances) which is a criterion for face validity (Brink, 2002:168).

Reliability
All data was gathered during the same time frame with all units commencing data collection during the same week.
All facilities had functional labour ward clocks which were used to note the times to ensure the reliability of the data.On the day of commencing data collection, the various facilities were visited to ensure that all clocks were in working order, as was confirmed the week prior to data collection.The re corded data could be verified against the labour ward registers in each facil ity thus adding to the reliability of the data, this was facilitated by using the last three/four digits o f the patient record number to check against the la bour ward register and to ensure that one entry per patient was made.Facili ties using the "Cradle" patient infor mation management system were ad vised to use the last four digits of the patient record.
Pilot study No problems were reported with the use of the instrument as a recording mecha nism.What came to light during the pilot study was that on one occasion an ambulance arrived to transfer a pa tient, but then was unable to take the patient because the condition o f the patient did not allow the EMS crew to "load" the patient as other, non-obstetric patients in the ambulance were needing to be transferred to facilities other than the referral hospital of the unit.The implication of this situation was that waiting time was increased because another vehicle would need to be dispatched to transfer the patient.
The decision was made to include these situations when recording the data and to note the time of the first call and the time of arrival of the ambulance that takes the patient to the referral hospi tal.During the pilot study a further ex clusion criterion was identified; pa tients requiring transfer for tubal ligation were excluded from the sample as these did not require definitive in tervention influencing morbidity and mortality.Data collection for the pilot study took three days to complete af ter which the completed instrument was collected and minor changes made.

Ethical and legal considerations
No patient identifying data was col lected thereby conforming to the ethi cal principle o f confidentiality.No sig nificant time commitment was required from already overworked MOU staff in order to record the data to detract from patient care and other duties.The data collection period lasted less than one week and was therefore not signifi cantly inconvenient to staff.No risk to patients could be identified in the data collection process.A potential benefit to patients is that this study could in form health service policy makers and leaders around improving emergency transport capacity in the Cape Penin sula.A potential benefit for the service is that this study provides evidence of one of the problems experienced by the staff in primary maternity care.

Data collection
Coordinators managing the various MOUs were briefed by the researcher on the data collection procedure and the data collection instrument was dis tributed at this briefing.This informa tion was then relayed by the facility coordinators to the facility managers, who w ere then co n tacted telephonically by the researcher to re inforce the data collection procedure.Data collection was commenced simul taneously in September 2005 and this was ensured by a telephonic reminder by the researcher to all the units in cluded in the study.The availability of functioning clocks in the labour wards was confirmed with the respective fa cility managers prior to the commence ment of the study.
On the day o f the commencement of data collection, facilities were visited in the afternoon to follow up any prob lems or queries that may have existed.None were reported.The follow up visit also served as an opportunity to check that the clocks in all the facilities were in working order to ensure reliability of the data.
The data collection period ranged from two to five days at the various MOUs.There was no data collection over a weekend or during a week with a pub lic holiday as this was specifically ex cluded when choosing a week for the sample; these periods were assumed to place extra demands on the emer gency medical services due to in creases in alcohol-related motor vehi cle accidents and violence.Completed data collection instruments were then collected after being notified by the facility managers that they were com plete.The collected data could then be verified by correlating the information with the facility admission register.

Data analysis
The collected data was entered into an Excel® spreadsheet and the response times in minutes were calculated.Two data entries from two different units were excluded from the data analysis because one patient with a diagnosis of preterm labour had given birth at the facility prior to the arrival o f the ambu lance which was summoned 115 min utes earlier.The other excluded entry was for a call to transfer a patient with preterm rupture o f membranes, whose family decided to transport her them selves to the referral hospital 120 min utes after the ambulance was called.The sample size was therefore 48 calls, which met the required minimum of 43 responses.Mean, median and stand ard deviation of the response times per unit as well as the whole sample of 48 calls were determined.
The priority categories o f calls were analysed in the same way.The median was used in comparing the response times to the various categories o f calls and various MOUs because the distri bution was uneven, thereby giving a more accurate picture o f the data as opposed to using the mean (Brink, 2002: 185).The Kruskal-Wallis test was ap plied to determine the P-value to ascer tain whether there were any statistically significant differences in the median response times o f ambulances to the various units and various categories of calls.A P-value of < 0.05 would indi cate a statistically significant result.

Results
The response times of all the calls from the five MOUs included in the study are presented in table 1.The response times for all calls (n=48) had a mean of

Discussion
This study was planned and conducted under optimum conditions where there was no public holidays or weekends.Such conditions would place extra de mands on the emergency medical serv ices with increases in trauma related emergencies i.e. these results illustrate an optimum scenario.It would mean that during such periods the response times to calls from MOUs could be as sumed to increase due to the extra de mands on the EMS.The absence of statistical differences between the re sponse times of the various priorities of calls is cause for concern, because it can be inferred that there is no mean ingful benefit to the way calls are prioritised in the present system.Fly ing squad calls, which are requested for the transfer o f severely compro mised mothers and foetuses, appear to carry no advantage in the transfer of patients needing definitive urgent care.
The appropriate communication be tween control room and responding ambulances was brought into question during the pilot study, when an ambu lance that was "too full" to take a pa tient arrived at the MOU.This situa tion prolongs the waiting time to trans fer patients, as another am bulance needs to be despatched to ultimately transfer the patient.

Limitations of study
Although the study sample was small, it was calculated on the basis of 90% power and was conducted over a short period of time under what was consid ered to be 'ideal' circumstances where no extra demands are made on the EMS in terms o f weekends and public holi days.
There was dependence on the staff working in the MOUs for the collec tion of data which is considered a pro cedural limitation.The reliability of the data from one facility was questioned because there were no data recordings on the instrument for a period of 18 hours despite there having been trans fers (two for the period) noted in the labour ward admission register.The recorded data on the instrument was however used in the analysis.The 'missed' entries were not considered to be significantly influential on the findings because although two con secutive calls were omitted the reasons for the calls and time frame of responses did not appear very different to those recorded on the data instrument.
Flying squad calls have an added time lag because these calls are placed via the Groote Schuur labour ward by the senior obstetric registrar "on-call" and not directly from the MOU clinician.
Somerset and Mowbray M aternity Hospitals also place flying squad calls from their respective MOUs through this route thereby adding time to the response time for this category of call if not done immediately.This may have influenced the recorded response times for this category of call because up to three people would be involved in mak ing three calls to transfer a single pa tient.This chain of calls may influence the ultimate time it takes an ambulance to respond to this category of call.The study did not attempt to measure this pattern and its influence on the find ings.
This study did not look at the causes of quick or delayed responses, neither did it report on the clinical outcome of the patients who were transferred, as these were beyond the scope o f this study.

Recommendations
The system o f call prioritisation in the PMNS and EMS at the time that this study was conducted needs revision.
The system of prioritisation of the vari ous categories o f calls does not appear to be the same in the PMNS and EMS.
Call prioritisation in the EMS was pre viously "allocated by the caller" (Dr Wayne Smith, 2005, personal commu nication).The need for having flying squad calls placed via the Groote Schuur Hospital labour ward added to the time delay when needing to have patients transferred for urgent, higher levels o f care.This should be reas sessed when revising the policy, so that the placing calls for this category may be undertaken by the medical staff of the secondary re ferral centres The feasibility and availability o f us ing community members or private con tractors to transport non-emergencies to higher levels of care may be worth while investigating.Diagnoses such as prolonged pre-labour rupture of mem branes and neonatal jaundice may be considered to be transported in this way.This situation could indicate that this is not an appropriate use of the EMS, where private contractors could be used to transfer patients not requir ing medical monitoring en route.This may relieve the patient load on the EMS and allow more effective use of mate rial and human resources.In some cases it may be feasible and appropriate to use the families o f patients who have vehicles for the transfer o f stable pa tients to higher levels of care.These kinds of transfers can be considered on an individual basis depending on the severity of the problem and discus sion between relevant referring and accepting clinicians.
In order to improve the situation as il lustrated in the findings of this study, public-private partnerships with private sector emergency services should be investigated.The circumstances under which a private ambulance could be summoned e.g. in terms o f time delay should be determined.This should be done after careful evaluation of the cost implications for the provincial health department in these situations.
1 Information elicited from EMS staff member on condition o f anonymity.
It may be o f value to conduct a larger scale study over a longer period of time and to investigate the implications and consequences of delayed or quick re sponse times on maternal and neonatal morbidity and mortality in the PMNS.It may also be worthwhile to conduct a comparison with other regions and provinces, providing an opportunity for further, more extensive research on this topic.

Conclusion
A m bulance response tim es in the PMNS are cause for concern as proven in this small but significant study.It confirms the perception of delays ex perienced in the MOUs o f the PMNS.The Western Cape is considered to be better resourced than other provinces and yet the time it takes to transfer maternal and neonatal emergencies leaves considerable room for improve ment.

(
Mowbray Maternity and Somerset Hospitals).The present list of diag noses that constitute 'flying squad' emergencies can be used as a guide by the staff at secondary maternity facili ties.Since this study was conducted the call prioritisation criteria have been amended and were introduced in the EMS and PMNS in April 2006.No link could however be established between the findings of this study and the revi sion o f the transfer protocols.If delays are due to capacity or resource constraints, community mobilisation and public/private partnerships could be considered.The communities served by the MOUs all have established transport infrastructures such as roads.