How fast is fast enough? An audit and league table of response times of acute hospital NHS Trust switchboards in England

Abstract

Summary

Objective

To audit response times by all acute NHS hospital switchboards in England for external incoming calls. As with any audit, we set out to draw attention to efficiencies and deficiencies in the system, highlight problems, and suggest the first steps towards improving standards in a very important and poorly acknowledged part of our healthcare system.

Design

Telephone calls were made to hospital switchboards, and time to response was measured. The results were then audited.

Setting

Acute hospital NHS Trust switchboards in England.

Main outcome measures

Time to response by hospital switchboards. The target time was 20 seconds.

Results

Only 36% of 219 hospital switchboards met the 20 second target, with the average answering time across all switchboards being 45 seconds. The best performer was University Hospital North Durham (mean = 1 second) and the worst Bristol Royal Infirmary (mean = 381 seconds). We found that automated answer systems substantially increase the answer time to an operator; the fastest automated service was at Princess Royal University Hospital, which was ranked 69th.

Conclusions

We recommend a formal NHS-wide audit system for switchboard response times.

Background

The hospital switchboard lies at the heart of the NHS, playing a critical role without which the NHS would come to a standstill. As well as linking professionals to each other to allow proper running of patient services, linking professionals to patients and linking patients to relatives, it also handles emergencies and major incidents, as well as running the bleep system and handling on-call rosters. Response time remains the key part of the standard of service that a switchboard provides; long response times have the potential to reduce the general efficiency of a hospital as well as increase frustration not only for staff, patients and relatives but also for pressurized switchboard staff themselves.

Casual enquiries into the running of switchboard systems in a number of hospitals showed the absence of an NHS-wide or published audit system for reviewing hospital switchboard performance. No formal guidelines could be found within the NHS for what constitutes an acceptable response time, although recommendations from outside the NHS do exist (

Table 1

Recommended maximum response times for telephone operator services outside the NHS

Operator service	Website	Response time
City of London Police	www.cityoflondon.police.uk	90% within 20 seconds
Borough of Rochdale	www.rochdale.gov.uk	7 rings or 20 seconds
University of Salford	www.isd.salford.ac.uk	Within 15 seconds

Table 1), as do independent guidelines within individual NHS trusts.

Interestingly, the telecommunications regulator Ofcom could not provide us with information regarding guidelines or standards for call answering times either from their enquiry line or from their website, and such information could not be found on extensive searches of the following websites:

National Audit Office (www.nao.org.uk)

British Telecom (www.bt.com)

Ofcom (formerly Oftel) (www.ofcom.org.uk)

Health Care Commission (www.healthcarecommission.org.uk)

Google (www.google.com)

We undertook an audit of all acute NHS trusts in England to obtain answering time information and produce a performance league table. We also provide recommendations for maximum response times based on our findings.

Methods

There are more than 1000 NHS hospitals in the UK. In order to audit a manageable yet meaningful number of hospitals, we chose to look at all acute NHS Trusts (i.e. those with Emergency Departments) in England. A complete list of acute hospital NHS Trusts was obtained from the Healthcare Commission website and the hospital telephone numbers obtained from the Department of Health website (www.doh.gov.uk).

In a one-month period, on three specific and different weekdays and at three specific timeslots (

Table 2

dates and times of audit data collection

Call	Date and time
1	Tuesday 18 April 2006, 10:00–12:00
2	Thursday 20 April 2006, 15:00–17:00
3	Wednesday 10 May 2006, 14:30–16:30

Table 2), all hospital switchboards from this list were telephoned from a private (non-hospital) line at the cost of the authors (657 calls). The prime objective was to speak to a human operator on all occasions. Data collection depended on the answer mechanism employed by the hospital. This could be either a direct human answer, through an answering system, or through a hybrid.

A stop-watch was used to record all times and times rounded to the nearest whole second.

Data collection

The flowchart in

Figure 1

Flowchart of the data collection process

Figure 1 describes the data-collection process. When the operator was obtained, the response ‘sorry, wrong number’ was given before hanging up (maximum time live with operator = 5 seconds).

‘Time-out’ or engaged tones were documented and the numbers re-tried until eventual success.

Results

A league table presenting the mean and standard deviation of response times of acute NHS hospital switchboards in England is available in Appendix 1. Ranking was undertaken first in order of response time (low mean) and next in order of consistency of response time (low standard deviation), then with preference to the single fastest response time. Those still not differentiated using these three rules were given an equal placing.

We found a vast discrepancy in actual answering times and the consistency of answering times throughout hospitals. The average answering time was 45 seconds. The best performing switchboard was University Hospital of North Durham and the worst Bristol Royal Infirmary. Selly Oak hospital also performed poorly, being not only near the bottom of the league table, but also remarkably consistent in its slow answering times (standard deviation of only 6).

Five switchboards gave engaged tones, hung up or ‘timed out’ at some stage (asterisked in the appendix). 71 (32%) of switchboards use an automated answering system, which added a mean of 23 seconds (range 19–65 seconds) to the answering time from an operator, due mainly to the length of the recorded message – resulting in added cost to callers. The best performing automated answer switchboard (Princess Royal University Hospital, Bromley) only came 69^th in the league table, and was the only such switchboard to answer within the 20 second target. The top 10 automated answer switchboards are shown in

Table 3

Top ten automated answer switchboards

Automated answer hospital switchboard	Time to operator (mean sec)	Overall ranking
1	Princess Royal University	19	69th
2	Grantham & District	26	93rd
3	Bishop Aukland General	26	95th
4	East Surrey	29	100th
5	Salisbury District	30	103rd
6	Bedford Hospital South Wing	32	110th
7	Royal Free	33	111th
8	Russell's Hall	34	113th
9	St Richard's	34	114th
10	Darent Valley	34	115th

Table 3.

Discussion

We set out to make some sense out of the state of NHS telephony. Our first finding was a lack of any national NHS standards for call answer times, so we set a standard answering time target of 20 seconds to reflect other guidelines for similar services that we could find, such as the police (Table 1). Response times from the hospitals in the study were then interpreted in this context.

We found that automated answering systems guarantee long delays, and thus significant added expense to callers, without major gain. No automated systems were in the top 50 for response time. If automated answering is to be used, there needs to be an evidence base justifying the additional cost and delay to callers. Where it is used, the potential opportunity for health promotion appears wasted.

The involvement of a private company in the operation of the switchboard with stringent internal audit processes and accountability appeared to be an important factor in making Durham the ‘best’ switchboard, as explained by them when questioned.

Our prime objective was to highlight the importance of hospital switchboards and their performance. Good performance requires setting standards and then striving to achieve them: unfortunately, this does not appear to be happening in NHS telecommunications at the moment. Some systems are in place that allow direct access, for example to General Practitioners, using special phone numbers. Other ‘hash codes’ exist for priority and fast-track lines but these are non-uniform, sporadic and often poorly utilized due to lack of awareness. The hospital switchboard is the ‘face’ of the NHS and a gateway for patients and doctors alike. We think that this study highlights a need for a national audit system to ensure a good quality service.

Box 1

Making the most of your hospital switchboard

Getting through to a switchboard operator can be quite a time consuming process. Here are some simple tips that can not only reduce the time you spend contacting switchboard but also their workload:

Ring if possible at off-peak times. 9–11 am and 2–4 pm tend to be the busiest times of the day.

Patients and relatives should be given clear details of the name of the ward or department to contact and their direct number so that they bypass switchboard.

Plan ahead – get all numbers that you are likely to need throughout that day in one call.

Use the hospital directory – copies are available on the wards and in secretaries' offices, theatres etc. Wards also often have a list of frequently used numbers at the nurses' station. Some hospitals provide a mini-directory, so if you can, carry it!

The hospital intranet system is another useful way to access contact numbers.

Keep a record of numbers that you commonly use either on your PDA or folder.

Ask your hospital to prepare printed cards with a list of regularly used extension numbers that are attachable to bleeps and give them to all new junior staff at induction. Your postgraduate centre can help you here.

Keep up-to-date rosters for your department, as these usually have a record of who is on call that week and their bleep numbers.

Some hospitals have a voice-activated system accessible via a separate dialling number. This enables you to get most numbers without having to speak to an operator. If you get on well with this system and it recognizes your accent, use it!

A prefix can be used (entered before the extension number) to gain direct access to most wards and departments when calling from outside the hospital.

When expecting outside calls, give extension numbers at which you can be reached that day; where permissible, use mobile phones rather than the bleep system. Note that some hospitals are mobile-only now.

We did not have the resources to audit all hospitals in the UK, yet we managed to assess a meaningful and representative sample of major hospitals and bring important issues to the fore. We could have changed our methodology, for example by providing median instead of mean values, but that could have disregarded single long answer-times and we felt that these should incur a penalty in league table position.

We hope that policy makers can tackle the issue of NHS telephony in light of the implications of this study for both patient care, efficiency and cost. A large-scale audit or re-audit would only be feasible with major financial backing and negotiations with telecommunications providers to implement local systems.

Conclusions and recommendations

Many hospitals (about one third) do provide an excellent service with fast response times to a live operator. The best performing acute hospital NHS switchboard in England for response times is the University Hospital of North Durham. Well done.

Recommendations

• We believe a ‘live’ operator should answer all calls to any hospital within 20 seconds; adequate local funding should be available to ensure that all Trusts can comply with this target for both internal and external calls.

• The NHS should introduce a formal, system-wide audit for switchboard response times, publishing its findings on a regular basis and covering all hospitals. Telecommunications providers can easily provide detailed analyses of call volumes and answer times and these are already used by some Trusts on an individual basis. This could be extended eventually to cover GP practices and private hospitals.

• A valuable first step in improving the standards of poorly performing hospitals would be discussions between the switchboard supervisors and managers of these Trusts and those performing well.

• All users of hospital switchboard services should endeavour to maximize switchboard efficiency through simple measures that avoid excess or unnecessary calls to the switchboard (Box 1).

• Hospitals currently employing automated answering could, and perhaps should, use the opportunity for health promotion whilst the call is waiting. However, when many trusts are able to provide a rapid and efficient service direct to an operator without an automated response system, we believe that such systems are redundant and detrimental in terms of call answering times and increasing the costs of calls.

• An engaged tone or time-out is unacceptable for an emergency service.

Footnotes

DECLARATIONS

Competing interests The authors are not affiliated to any hospital or NHS trust with regards to this study. KS has previously worked at the Bristol Royal Infirmary and soon after completion of this study ST commenced work at Durham Hospital. The authors have no other conflicting affiliations with hospitals, hospital switchboards or the telecommunications industry

Funding The authors have undertaken this study as general members of the public, outside of their NHS commitments, in their own time and at their own cost

Ethical approval Not applicable

Guarantor KS

Contributorship All authors undertook calls to switchboards, KS designed the study and wrote the manuscript with contributions from ST. The authors have rewarded the staff of the best performing switchboard with a prize and certificate at their own cost