The aim of the cervical screening programme is to reduce the incidence of, and morbidity and mortality from, invasive cervical cancer. This is achieved by screening eligible women between the ages of 25 and 65 years.
In my role as an acting Senior Biomedical Scientist in the cytology department, I screen and report negative and inadequate cervical samples in line with XXX protocols as well as facilitating the flow of abnormal cervical samples through to consultant pathologists and ensuring appropriate recall management is given to patients. I am also involved with the supervision of staff and the production and management of in-house standard operating procedures.
Description of Problem
Although always monitored by XXX, the timeframe from taking a cervical sample to when it was reported has never been given a strict turnaround time target, unlike the screening programme in England, which since 2010 has had a 14-day turnaround time target. However, this is to change as XXX want women screened in XXX to have an equitable service to that of the women in England and with the introduction of new testing alongside the current cervical cytology test, laboratory turnaround times need to be under 3 weeks.
In May 2013, the majority of our work was turned-around in 3.5 weeks with only 0.9% of our workload being reported within the proposed 3 week timeframe. Being aware that XXX were likely to implement changes within the next 12 months, we wanted to be proactive and improve our performance before XXX enforce improvement methods upon us or worse the service was moved from XXX.
Therefore, the problem was identified as: How can we reduce our turnaround times to under 3 weeks (from date received in laboratory to date authorised)?
Analysis of the Problem
We were aware as a service that our turnaround times had increased over the last couple of years due to the loss of staff through retirement and centralisation and due to the transportation of work to an off-site laboratory for processing and then its transport to our department for screening and reporting. However, to identify the areas that we could improve, we created a work-stream to understand the flow of work through the laboratory process.
A simplified version of the work-stream is found below:
(Removed for confidentiality reasons)
The days in the green ovals are average accumulative working days it would take to achieve a reported sample from arrival at the laboratory. These are working days and do not include weekends, therefore it is possible a further 2 to 4 days could be added to the actual turnaround time, making a possible estimated laboratory turnaround time of between 4 and 10 days. These days are calculated with no backlog, and therefore the next step would be to identify were backlogs or bottlenecks were occurring.
The processing of work at XXX was beyond our control. As was the transport to and from XXX, which occurs on alternative days and is governed by XXX. Therefore, the two areas we could concentrate on were the booking in and the screening of slides.
In May 2013, the administration staff based in XXX had a slight backlog of samples to book in, this equated to 1 to 2 days delay. The main cause of this delay was lack of staff and it wasn’t something we could address immediately as recruitment of new staff would take time and screening staff could not help as the backlog and delay in slides being screened was even greater. In the same month, the backlog of slides waiting to be screened was estimated at 2 to 3 weeks, this equated to around 3000 slides waiting to be screened.
The backlog of screening work was not only affecting swift results to the patients it was also having an effect upon the screening staff morale as they had to see large numbers of slides waiting to be screened, which could affect the turnaround further. We were also notified at this time that a member of the screening team would be leaving and therefore the urgency in reducing our turnaround times became more imminent as we would have less staff in a month or two. If no action was taken now the turnaround times would increase further once the member of staff left.
As a result, the area we need to concentrate on and identify possible solutions to was the matter of reducing screening times of slides, by attacking this element of the laboratory process we should be able to reduce our overall laboratory turnaround times.
Identification of Possible Solutions
The Cytology senior staff members, including myself, started with a brain-storming exercise of trying to identify possible solutions. Due to constraints and safe-practices put in place by the NHS Cervical Screening Programme (NHSCSP) and XXX it is not possible to ask screening staff to screen each slide faster, as false positives and false negative results would increase and the quality would be affected. And we could not ask staff to screen for a longer length of time in the day, as screeners can only primary screen slides for a maximum of 5 hours a day, again due to the quality risk. However, staff could work overtime by coming in at weekends.
To try and improve the workflow pathway within the screening room I decided to re-read the ”Cytology improvement guide – achieving a 14 day turnaround time in cytology” produced by the NHSCSP when England was undergoing changes to its screening programme turnaround times. This guide is a collection of evidence-based ideas for improving workflow and turnaround times in the laboratory and administration departments of the cervical programme.
We had already adopted one suggestion from the guide in 2010 regarding batching sizes of work, whereby we have reduced the number of slides in a tray from 20 to 8 slides. Although each slide was still being screened for the same time length, a tray of slides were being finished quicker and therefore the screening staff felt happier and kept the work flowing through the laboratory. However, on inspection the work wasn’t flowing, as after primary screening, trays were being placed into the rapid review box and only taken out for review twice a day (first thing in morning and after lunch), with some staff only reviewing slides once per day. This was causing the first bottleneck and it was also having a knock-on effect on the authorising of samples by senior staff as large chunks of authorising was occurring first thing in the morning or after lunch, which due to the monotony of checking details and results was causing errors to occur in reporting. We therefore needed to undertake a Lean transformation.
Lean is an improvement approach to enhance flow and eliminate waste that was developed by Toyota. Lean is basically about getting the right things to the right place, at the right time, in the right quantities, while minimising waste and being flexible and open to change. Lean thinking focuses on what the customer values, any activity that is not valued is waste. If you remove the waste, the customer receives a more value-added service. In our department this means reducing our turnaround times, so that the customer (XXX women) get their cervical results in 3 weeks. We therefore had to remove processes that were not adding value to the end result. By investigating further it was noted that some screening staff were writing on the front and the back of the sample request forms if the samples were negative. There was no benefit in duplicating the result on the back of the form and therefore as part of Lean it is considered waste.
Distraction was also a large issue in the screening room. To screen and report slides, it requires concentration to process and distinguish abnormal cells from normal samples. However, staff were regularly stopping screening to chat or employees from other departments were coming in to talk, make a cup of tea or to eat biscuits causing distractions of up to 30 minutes at times. The telephone was also a regular distraction, with service users phoning up for advice and results and the XXX administration department phoning up regularly to change and update patient managements.
Summary of Options
We eventually decided the options available to us were:
1. Screening staff to work overtime.
If every member of the screening team and senior team (Senior Biomedical Scientist and Cytology Manager) were to partake in overtime work on one day of the weekend and managed to screen an average of 30 slides per day the backlog could be cleared in 9 to 10 weekends. However, partaking in overtime is a voluntary decision and not all staff members would want to or be able to work overtime. This means that if we were to only adopt overtime as an option a minimum of 9 weekends would need to be worked, and as we were not changing the day-to-day workflow process a further period of overtime may be needed in the future to minimise any further backlogs. Working overtime also had a cost implementation, as staff would need reimbursing for their time given. As overtime would be over a weekend this would be calculated at time and a half pay.
2. One tray of primary screening followed by one tray of rapid reviews. To overcome the current bottleneck of work after primary screening, it would be more advantageous to rapid review slides little and often instead of first thing in the morning or after lunch. The preferred scenario would be to primary screen a tray of slides and then screen a tray of rapid review slides and alternate like this throughout the day. This means the throughput of work is more constant and turnaround is less. These examples explain this option further:
Current example –
As you can see this example of current working practice means the slides take approximately 22 hours from being initially primary screened to being authorised.
Proposed example –
This example shows that by alternating between primary and rapid review screening the whole process up to authorising should take approximately 75 minutes. A saving in time of 20 hours and 45 minutes for those 8 slides to be reported and authorised. Some work primary screened at the end of the day will have to wait until the next day for rapid review, but on the whole most
work will be turned around at a much faster pace. This could be implemented immediately with no cost implications or changes to staffs working pattern.
3. Authorising work little and often.
The examples used above also outline how this can improve the bottleneck between rapid review and authorising. Currently authorising is carried out in large batches after rapid review has been completed mainly at lunchtime and occasional first thing in the morning. These large batches of work are monotonous and transcription errors can arise due to this. By having rapid review completed in small batches of 8, it is possible for senior staff to authorise in these small batches as soon as rapid reviews have been completed. This means authorisation can be achieved much sooner and hopeful reduce the risk of mistakes occurring.
4. Reduce waste.
A Lean approach was considered and any process that didn’t add value was removed. The first wasteful process identified was writing on both front and back of request forms, writing on the back was considered worthless. This process approximately took 10 seconds to carry out. Therefore if this was stopped each screening staff member screening 30 slides a day on average would save 300 seconds a day. This equates to 5 minutes, which is more than enough time to screen one or two extra slides a day.
Which over a week could mean 10 extra slides been screened by one screening staff member or 80 slides a week if all screening staff stopped. The second process considered was the taking of primary screening work from the work pile before necessary. Some individuals would pick up 2 trays of primary screening or would leave unfinished trays on their desks overnight. If these trays were left in the pile or half trays placed back into the work pile before going home, it would allow another member of staff to screen them sooner and therefore have them moving along the work-stream quicker.
5. Adopt quiet-time.
The most controversial option considered was to implement quiet-time in the screening rooms. An initial quiet-time of between 9.30am-12.30pm and 1.30-3.30pm was suggested, asking all staff member to keep any talking to work-related questions only and to deter any colleagues from other departments from coming into the screening room or phoning the department. No mobile phones would allowed in the department during these times and work emails were to be checked first thing in the morning or first thing after lunch.
Evaluation of Options
The NHSCSP advise on using a PDSA cycle (plan, do, study, act) to evaluate and to test out ideas to ensure changes make the improvement required before implementation. Although the cycle is good at temporarily trialling a change and assessing its impact, it’s not useful in highlighting the best option or options to adopt without trialling them all. We therefore need to analyse the strength and weakness of each option first.
A decision matrix was created to evaluate and prioritise the 5 options above. The criteria we used to make the decisions were: Cost implications; would there be any extra cost involved with each option and was this a small or large cost. (Rated from 1 (high cost) to 3 (low cost)) Speed of implication; were we able to implement this option immediately or would it require much planning. (Rated from 1 (slow) to 3 (fast))
Effect on staff; would the staff be happy with these measure, would it cause a negative effect in the work place. Did the staff have to make large adjustments to their working practice. (Rated from 1 (high detriment) to 3 (low)) Speed of improvements; how long would it take for us to see a change in turnaround times if option was implemented. (Rated from 1 (slow) to 3 (fast))
Each option was rated for each criterion, so that we could establish any issues with each option.
From looking at the table above, options 2 and 3 have scored well on all 4 criteria and therefore would be able to be implemented with little disruption but with a good impact on turnaround times. Option 1 unfortunately had too many issues regarding cost and the time it would take to gain agreement for funding the overtime and for staff to agree to giving up their free weekends. Also every staff member wasn’t able to participate due to personal commitments and families, this meant the overtime worked could vary greatly week to week. It was therefore decided options 2 and 3 would be trialled further and option 1 would be abandoned.
Although option 4 would take time to see results, it was decided that by ‘Leaning’ up the work process could only be a good thing and that the ideas suggested to reduce waste in the workplace should be implemented immediately.
The main cause for concern with option 5 was the detriment to the staff, however this change to working practice was considered to be the one that would have the greatest affect on reduction of turnaround times. It was therefore decided to trial this for a while, and that once the backlog was reduced we could relax the quiet time option and re-implement it if the backlog was to increase again.
We could now look at the PDSA cycle and using turnaround time as a measurement assess the capabilities of option 2,3,4 and 5. By starting to trial the options and regularly monitoring and studying the turnaround times (Do & Study) we can highlight any changes that may be necessary (Act) and think of new plans to tackle any issues that arise (Plan).
Before implementing and disseminating the information to all the staff, the cytology manager and cytopathologist went over all the presented options and made the final decision to implement the improvements for reducing turnaround times. They were happy to put forward the 4 options identified to all the staff and create action plans for implementation.
A staff meeting was held to explain the situation regarding increasing turnaround times and the options put forward in the form of an action plan to the staff on how we could reduce the turnaround times. The action plans presented were:
It was important to us to use SMART action plans, which meant we had to be specific in our actions and create measurable, achievable and realistic plans. An achievable time frame was set by creating weekly and monthly milestones for reviewing our turnaround times.
Staff were asked to provide their thoughts on these actions and identify any issues or bring forward any options not already covered. Staff members made known their fears regarding quiet time, with one staff pointing out their working times were 9.30am until 2.30pm, and therefore apart from lunchtime she cannot talk during her working hours. This was taken into consideration and the hours of quiet-time were altered. It was also stated the quiet time was a short term proposal and would be lifted as the backlog reduced/ended.
An idea of moving the boxes holding the work needing rapid review and the boxes holding forms for authorising was also put forward. As currently the boxes were located across two different rooms and in the furthest corner of one room, meaning the entire screening room was being disturbed when someone went to put work/forms in the boxes. By moving the boxes to an equidistant place between both room, it would reduce the time wasted in walking to the boxes and reduce disturbance to other staff. Therefore the new and modified action plans were added.
The staff was also advised that the turnaround times would initially be monitored in-house on a weekly basis to trial these changes and if the turnaround times were improving the monitoring would default to the XXX monthly monitoring. The aim was to try and authorise 300 cases a day, this was a realistic number to screen between 8 screening and 2 senior biomedical scientist staff. This equated to 1500 slides a week authorised, and as approximately 1000 slides were received and booked in each week, this meant we would be clearing approximately 500 slides from the backlog each week.
Review of Outcomes
Work levels were reviewed weekly by the Cytology Manager using the laboratories computer system. It was possible to see that the workload being reported was greater than that of the workload coming into the laboratory, and therefore the backlog of work was decreasing. However, what helped staff morale and made them aware of a reduction in the backlog and an improvement in turnaround times was being able to visualise the time between the date the sample was taken and the date the sample was screened and also seeing the reduction in the pile of work that was waiting for screening. As all options were implemented at the same time it was not possible to identify the effects of each individual option.
XXX were also gathering the 3 week turnaround figures for all laboratories and these were being disseminated amongst the screening staff on a monthly basis. From the graph provided by XXX, it is clear from the implementation of changes in June 2013 the percentage of samples turned around in less than 3 weeks has steadily increased.
A slight dip in the percentage is noted between August and September 2013, this coincided with the loss of a staff member and a move to a new laboratory building, which resulted in screening down-time for a couple of working days. It was at this time a decision was taken to stop the strict enforcement of quiet time, allowing the occasional ‘chatter’ within the screening room. This made work more welcoming and created a happier work environment, this in turn seems to have helped in the increase of samples turned around in 3 weeks.
The introduction of these measures over the past 12 months have seen the departments 3 week turnaround times increase from 0.9% to 100%.
NHS Cervical Screening Programme (NHSCSP) – Cytology improvement guide: achieving a 14 day turnaround time in cytology
NHS Cervical Screening Programme (NHSCSP) – Continuous improvement in cytology: sustaining and accelerating improvement