Knowledge Matters Volume 8 Issue 4

issuu.com/SECQO

twitter.com/SECSHAQO

www.QualityObservatory.nhs.uk http://www.networks.nhs.uk/nhs-networks/sec-qo

Volume 8 Issue 4 October 2014

This year is the 75th anniversary of the film of The Wizard of Oz*, so this issue we’re going to follow the yellow brick road to find out all about the UK Improvement Alliance, the NHS Safety Thermometer event, and population segmentation.

There is however a parallel to be made between Dorothy’s travelling companions and the NHS! The lion, the tin man, and the scarecrow wanted three things from the wizard: courage, heart, and brains. These three things, I would suggest, are what any NHS employee needs to make the system run smoothly and efficiently. Yeah—I know that sounds obvious, but how many ideas have we all seen around the system, that are, seemingly ill-thought out, or ones that seem intelligent but actually don’t take caring for the patient into account, or those ones that are just too timid and don’t go far enough because people are afraid of the consequences. I’ve been in the NHS long enough to have seen all kinds of this stuff floating around. It’s difficult to get the balance right. Cold hard facts can tell you that the sensible thing to do for the system is to shut a hospital, but then that’s perceived as heartless to the patients using that hospital, and so a muddled compromise is made that, in the end, increases inefficiency and patient dissatisfaction. So I can only hope that we all use these three qualities in a balanced and thoughtful way, because if the NHS is to thrive through change this is the best way forward.

And with that I’ll click my heels and say “There’s no place like home6”

*It’s also the 75th anniversary of Batman this year, and in hindsight if I’d thought of that sooner I’d have been photoshopped into a cooler (although probably no less camp) picture!

Welcome to Knowledge Matters

Inside This Issue :

Population segmentation using adjusted

clinical groups

2 National Cancer Experience Patient

Survey

8 News 11

Questionnaire design in health research 4 Has the Friends and Family Test made a

difference?

9 Back Page 12

Safety in Numbers: NHS Safety Thermom-

eter Event

6 UK Improvement Alliance Launch 10

2

[email protected] www.QualityObservatory.nhs.uk

In recent months there has been an increased interest in population-level healthcare utilisation analysis

using an approach known as segmentation. With this approach, rather than seek to analyse healthcare data

by activity type (e.g. emergency admissions, outpatient appointments and so on), we seek to group everyone

in the population into one of a number of distinct segments (case-mix categories) based on features that

mean they are likely to consume similar volumes of healthcare – and then understand utilisation through that

framework. Datasets like those underpinning the ACG system facilitate a comprehensive approach to such

methods because unlike most other healthcare datasets, ACG data contains the necessary demographic

information for everyone in the population who is registered with a GP - regardless of whether or not they

have come into contact with healthcare services.

National initiatives such as the Better Care Fund have reinforced the need for Clinical Commissioners to

segment the populations they serve in order to identify the segments of the population who utilise

disproportionate volumes of care and to help target specific projects such as Integrated Care. There are

many ways that a population can be segmented; for instance by prior utilisation, predicted risk of future

events such as emergency admission, age, specific diseases and so on. With any approach, the principle of

mutual exclusivity between segments is important – if people could belong to more than one segment, it

becomes much harder to describe the breakdown of healthcare utilisation across the population, due to

some volumes being accounted for in more than one place. This presents a problem in approaches seeking

to use specific diseases for segmentation, because any individual may have between none, and many

diseases – and could therefore belong to several segments.

Two potential ways to utilise the obvious value of disease information but avoid the double-counting issue

are to either; 1) allocate individuals to disease-based segments on a hierarchical basis i.e. if they are already

allocated to one group, they are excluded from all others, or; 2) allocate individuals to segments based on

the total number of diseases they have. We have adopted the latter approach to date in Central Southern,

following other work using ACG data that has identified that it is the presence of multiple diseases, also

known as multi-morbidity, that tends to drive high healthcare utilisation.

In order to use a subset of diseases that are well-defined and likely to be uniformly well recorded across the

board, the diseases falling under the Quality and Outcomes Framework (QOF) Long Term Condition (LTC)

disease registers were used. This definition of LTCs

constitutes eighteen specific diseases; Asthma, Atrial

Fibrillation, Cancer, Chronic Kidney Disease, COPD,

Coronary Heart Disease, Dementia, Diabetes Mellitus,

Epilepsy, Heart Failure, Hypertension, Learning Disability,

Mental Health, Obesity, Osteoporosis, Peripheral Arterial

Disease, Rheumatoid Arthritis, and Stroke. Initial analysis

for one CCG identified the pattern of multi-morbidity shown

in the chart on the right;

As we might expect, most people in the population don’t have a QOF LTC and as we move through levels of

multi-morbidity there are decreasing percentages of the population in those groups. Next we cross-tabulated

the above groups with broad age groups representing young people, adults of working age, adults of

retirement age, and older people, which gave the following splits;

Population Segmentation using Adjusted Clinical Groups (ACG) data

By Chris Morris, Principal Analyst, Health Intelligence & Analytics Team, NHS Central Southern CSU

3


This table shows that

most people in this par-

ticular population were

both younger than retire-

ment age and tended to

have zero or only one

QOF LTC. The N/A’s

above are where we

found so few young peo-

ple with more than two

QOF LTCs that it made sense to roll them into a group described as 2+. Once the above groups had been

derived, we then used them as the framework through which to view healthcare utilisation. The table below

shows the disparities between the percentage of people in the segment and both the percentage of general

acute healthcare cost, and percentage of primary care activity that they account for. For example, the group

with 5+ QOF LTCs, aged 80+ makes up only 0.4% of the population, but accounts for 4.3% of the general

acute spend and 1.8% of primary care activity. A different way of making the comparison is to look at the dif-

ference in average general acute cost or primary care activity for a particular group, compared to other

groups; it is quite clear that the more multi-morbid segments associate with higher average healthcare utilisa-

tion.

Datasets that facilitate such analysis, and segmentation methods themselves are still in their infancy in the

healthcare context, but approaches like these are valuable, flexible, highly scale-able, and can inform intelli-

gent commissioning in ways that other methods can’t. It is this author’s belief that in addition to segmentation

methods such as these, which define a typology of the population, work should now also focus on defining

more general typologies of activity other than the standard, mandated typologies that already exist. This

would help us to understand where and why healthcare systems appear to be out of balance – for example,

with a generalised typology of admitting diagnosis where we differentiate between general acute diagnoses

that shouldn’t normally require admission, and more severe diagnoses such as those from LTCs, we may be

able to really illuminate not just where there is unwarranted variation, but really pinpoint why, in order to sup-

port truly evidence-based commissioning.

4


A couple of months have elapsed since I joined the Friends and Family Test team at the Quality

Observatory. Strangely enough, I have sat down to sum up what I have learnt and the major

challenges in my new job – I actually did not have exciting things to do on that Saturday evening!!

A notable challenge for me is to reply to queries received by email or phone calls on the Friends

and Family Test. One particular question was – “We need to roll out FFT to ask for our patient’s

feedback using a questionnaire. All I want to know is what exactly do I need to do for this

questionnaire?” My immediate reply was: Uhhhhh.. May I send you an email please??

Questionnaire design may seem straightforward but it actually requires considerable time and skill.

Why is that so? Wouldn’t it be right to say that it is, at the end of the day, just a matter of asking questions and getting

answers? Well, it becomes slightly intricate due to our wonderful brain which stores memories: Memories can be long-lasting

but they are also prone to inaccuracies and disremembering.

According to the cognitive model of mental processes, responding comes in 5 parts:

• Encoding in memory: respondents have to have some knowledge or memory of what is

being asked.

• Comprehension: respondents understand the question and relevant concepts.

• Retrieval: respondents retrieve the information from memory.

• Judgement: respondents assess completeness and relevance of what they remember.

• Communication/response: respondents decide whether their answer fits the response

categories and also decide whether they actually want to provide the answer or provide

one that might be socially acceptable.

Therefore, how well a questionnaire is designed will undeniably determine the quality of data obtained.

Obtaining quality data is particularly important in health related studies. Health research has incessantly influenced mortality

rates, communicated the practice of healthier lifestyles and driven people towards disease prevention activities. The main

challenges faced by health surveys are data reliability and accuracy, validity of responses obtained and representativeness

of samples. As much as the selection of the right methodological approaches is important, it is equally essential to provide

appropriate attention to the proper instruments used to collect data, one of which is the survey questionnaire.

In questionnaire design, our foremost challenge is the defining of the objectives of the questionnaire.

For example, let’s say Hospital X wants to conduct a survey to understand the satisfaction of

patients in Ward Y. What are its objectives? Does it simply want to know about overall satisfaction?

Or does it want to delve deeper and understand what influences their satisfaction level? What are

the areas of improvement? Etc.

Questionnaire Design in Health Research

By Trishna Julha, Programme Support Analyst

5


Questions should be generated to reflect the objectives of the questionnaire. A few points to

keep in mind when generating questions are:

• We may need more than one question to measure the same item in order to increase

reliability.

• Wording is crucial to obtain precise questions.

• Avoid double barrelled, ambiguous, biased or leading questions.

Oops .. I forgot one important stage: Before designing a questionnaire, the type/method of analysis that will be performed on

the information gained has to be considered. Both qualitative and quantitative techniques are available.

For example, it may be difficult or impossible to analyse the results of a questionnaire with open ended questions in a

quantifiable way: What did you eat yesterday? ‘Apples, Apple Tart, Apple compote, drank Apple Cider’. It is impossible to

know the number of apples you ate from this answer, if this is what we actually wanted to know!

Did I mention open ended questions there? There are two types of response formats for the questions in a questionnaire:

• Open-ended questions: There are no pre-determined responses, with more chance of

obtaining detailed responses. But less structured, difficult to encode and longer to complete.

• Closed-ended questions: Involves a list of response options (for example, yes or no or don’t

know). It takes less time to complete, but it has less depth and may lead to frustration if the

desired response does not fit options available. You may choose a several point response

scale, for e.g, the Likert scale.

Once the questions have been designed, the introduction has been added along with the confidentiality clause and attention

has been paid to ethical issues, we would think that we are good to go - a big fat no it is!

We should check if the questionnaire is measuring what was intended in a reliable and valid way. Remember the cognitive

model of responding, and our brain and its memories? We should therefore pilot test the questionnaire to verify that the

questions are working or measuring as intended, the extent to which it is consistent, and whether information that

respondents need to enable them to answer accurately, is available.

Questioning is an art indeed, to be attempted diligently. By creating and conducting health surveys that

are firmly based on quality techniques, researchers today can help form the outcomes of tomorrow.

Knowledge is having the right Answer and Intelligence is having the right Question!!

References

http://www.ons.gov.uk/ons/guide-method/method-quality/general-methodology/data-collection-methodology/services-available-from-dcm/cognitive-testing/index.html

http://www.uniteforsight.org/global-health-university/quality-survey

6


The ‘Safety in Numbers’ NHS Safety Thermometer summit took place in London on Thursday 23 October

with 200 delegates from 100 organisations taking part in the day. The event was a great success with Rt

Hon Jeremy Hunt launching the Patient Safety Briefing, the ‘next generation’ NHS Safety Thermometers

being released and the work done so far with the ‘classic’ NHS Safety Thermometer being celebrated.

I was very fortunate to be presenting some of the data from the Safety Thermometer ‘classic’ which has

been undertaken by over 5 million patients, an enormous achievement to be celebrated! My co-presenter

was Mr Dave Shackley, the Clinical Director for Urology at Salford Royal, who was particularly concerned

with the catheter data from the NHS Safety Thermometer. He gave a really compelling presentation, using

the data as part of a story of the need for change, reminding me how important context and narrative can be

in the communication of the wealth of rich data we have available to us.

Other presenters on the day also helped to underline the importance of measurement in achieving real

improvement for patients, and on the importance of individuals collecting harm data to better understand

and issues and target improvement. NHS England North Director of Nursing, Gill Harris gave an overview

of the Open and Honest Care initiative, which advocates the public sharing ts; Dr Mike Durkin, National

Domain Director of Patient Safety, NHS England, then took to the stage to of key patient safety and

experience measures. You can learn more about their work here: http://www.england.nhs.uk/ourwork/pe/

ohc/

The afternoon focused heavily on the ‘next generation’ of NHS Safety Thermometers with breakout sessions

for:

• Maternity

• Mental Health

• Medication Safety

• Children and Young People

Two films were shown at the beginning of sessions to

help highlight the importance of the ‘next generation’

work and what it is helping organisations to achieve.

I’d really recommend you take a look:

Maternity: http://vimeo.com/109022820

Mental health: http://vimeo.com/107608482

These films reminded us just how impactful good data and measurement can be; it’s not always seen as

sexy (at least not outside this office!) and is a means to an end, but without it, critical improvement just

wouldn't happen.

I covered the maternity session with a long term colleague from NHS England North, Debby Gould (pictured

looking like what I’m saying might be making some sense above) and, as ever, was struck by the

commitment of delegates to improving care for women and babies. There was also a clear understanding of

why collecting and analysing data is such a key part of this and a strong willingness to engage in the

Safety In Numbers: NHS Safety Thermometer Summit

By Kate Cheema, Head of Commissioner Development

7


measurement agenda.

Here at QO towers we’re particularly proud of our work on the ‘next

generation’ Safety Thermometers. Our crack development team have

delivered the online data collection mechanisms for all the next generation

thermometers (just look at those pretty colours!) as well as the analytics to

go with them. As I type, the team are working on the next stage, a cross

platform app for data input into the Safety Thermometer ‘classic’, enabling

users out in the community who may have no 3G or wifi access to input

their data on the go.

The day wasn’t all about the NHS Safety Thermometer (aww, really??)-

Ed.); the star of the show was the Rt Hon Jeremy Hunt who came along to

launch the patient safety briefing. Patients at NHS hospitals across the

country will soon be benefitting from ‘airline style’ safety advice about their

stay in hospital in an attempt to reduce avoidable complications – such as

blood clots, pressure ulcers, or falls.

Based on the concept of safety advice given on aeroplanes before they

take off, patients will be shown a film and provided with an information

card to read to

help them look

after

themselves

during their hospital stay. The safety advice

looks at simple things patients can do while in

hospital to make their stay safer. You can

access the video and a whole host of resources

here:

http://harmfreecare.org/

Patient+safety+briefing+film+launched+

If you want to see a quick 2 minute snapshot of the day (including a snippet of me waving my arms about)

the team have made a film you can see here: http://vimeo.com/110121115

It was fantastic to be involved in such a great day, celebrating success but equally looking to the future of

safety measurement. To finish I thought I’d share my personal highlight; at the first coffee break of the

morning, just after I’d done my turn, a senior nurse from an acute trust approached me clutching an iPad

(other tablet devices are available). She proceeded to show me the dashboards she’s developed using the

Safety Thermometer data, triangulated with other local information on key harms, asking my advice on

presentation and other possible visualisations. I was bowled over by her enthusiasm and commitment to

using the data with her teams on the frontline, to help them drive and take ownership of their own

improvement. Sometimes it feels that we analyst types labour in a vacuum and feed data into a black box,

but this shows how powerful our work can be and how we can always maintain a direct line of sight on

making things better for patients. Something to celebrate indeed!

8


National Cancer Patient Experience Survey 2014

By Rebecca Matthews, Specialist Information Analyst

The results of the 2014 National Cancer Patient Experience Survey have now been published (as of 25th

September) and the results are available to view here: http://www.quality-health.co.uk/resources/surveys/

national-cancer-experience-survey/2014-national-cancer-patient-experience-survey

This is the fourth National Cancer

Patient Experience Survey to be

published by Quality Health and

covers all adult patients in active

treatment for cancer in England,

defined as patients aged 16 or

over with a primary diagnosis of

cancer who had been admitted to

an NHS hospital as an inpatient or

as a day case patient and had

been discharged between 1st

September and 30th November

2013. The three month time

period matches that in the 2010, 2012 and 2013 surveys. The 2014 survey had 70,141 responses nationally

(from 153 NHS Trusts) and resources available on the website include a national report alongside individual

reports for each NHS Trust that returned data.

The whole of the trust level data is also available to download as an Excel file. All data is available by

tumour type and is split by the following cancer types – breast, lower GI, lung, prostate, brain/CNS,

gynaecological, haematological, head and neck, sarcoma, skin, upper GI, urological, and ‘other’.

Questions in the survey cover the whole of

the patient pathway, from a patient seeing

their GP to hospital care and home

support, so a huge amount of data is

available in the survey results, especially

as they can also be split by age and sex.

Year on year scores are also shown for

comparable questions. All of the reports

and data for 2014 and previous years are

available to download from the website

now.

9


So, has the Friends and Family Test made a difference? By Samantha Riley, Director of Insight, NHS England

I took up post at the start of 2013 as Director of Insight at NHS England

– the lucky person with responsibility for the seemingly

controversial Friends and Family Test (FFT). One of the things that

became clear to me very early on was that there were many strongly

held views about what FFT might achieve. It’s fair to say that most

views expressed were at best cynical. Some were strongly opposed. A

quick google on FFT will bring up headlines such as ‘not fit for purpose’,

‘not reliable’ and even one suggesting that FFT will ‘do more harm than

good’.

So, 18 months on from the implementation of FFT in inpatients and A&E

(and subsequently maternity services), has the Friends and Family Test

made a difference? Well, today NHS England released FFT results for

August – the 17th month of publication. I wonder how many of you are

aware of the extent of feedback already generated? To date just over 4 million responses have been

received – feedback on an unprecedented scale. A pretty impressive figure I hope you agree. How does

this compare to the annual inpatient survey? The survey that has been running since 2002. I haven’t come

across many staff working within the NHS who believe that this survey has made a difference so far. Well, it

receives approximately 65,000 responses each year – which to my reckoning means that we would wait for

over 60 years to receive as many responses. That’s a long time.

There are of course differences between the survey and FFT. The survey is statistically comparable and

FFT is not. Surely this is a good thing. Annual surveys and FFT are complementary, and different

tools. Organisations with embedded feedback loops who use FFT feedback will see their annual survey

results improve.

FFT provides near real time feedback on a continuous basis and there is increasing evidence that this is

having a positive impact on staff morale. Many front line staff are sharing similar experiences to a Senior

Sister at Leeds Teaching Hospitals who recently reported that: “It has been great to get weekly feedback

from our patients, which we can share with the staff. The vast majority of patient comments have been

incredibly positive, so it has been very motivational for the team.”

The most important component of FFT is that it provides actionable feedback – the free

text comments provide the rich information which describe the experience of patients

both good and bad. There are already many hundreds of examples where FFT

comments have led to improvements in patient care. A Chief Nurse and Director of

Clinical Standards who I met recently expressed the opinion that having worked in the

NHS for over 30 years, her view was that FFT had had the biggest single impact on

improving care.

On that note I rest my case. The evidence to me is clear. Readers can decide for themselves.

10


UK Improvement Alliance

On Wednesday the 29th of October the UK Improvement Alliance (UKIA) held a launch event. UKIA has set out to bring together quality improvement organisations to support each other and ‘go further, faster together’. A couple of team members have been involved in some early discussions, but this event gave some other team members to go along, meet some like minded people from other organisations and network a bit. Here’s what they thought:

“It was indeed a privilege to be among experienced and talented clinician/measurement leaders of the home na-tions of the United Kingdom to discuss the future ‘gets’ and ‘gives’ possibilities through a UKIA membership. What I took away with me – the motivation to reflect on healthcare quality improvement and the importance of network-ing to achieve this national goal; that I was witnessing history in the making as a knowledge base on healthcare and safety monitoring and improvement was being developed; that measurement or any of my daily tasks are slightly scary to many and the Alliance will enable collaboration to create a better understanding; a further appreci-ation of the work accomplished through Friends and Family Test (my main task!) to improve healthcare; and how bizarre I must be because everyone else present seemed to have pets and/or children!”

-Trishna Julha, Programme Support Analyst

“I was unsure what to expect from the day when I first arrived at the UKIA launch event in London. I hadn't previ-ously been involved in this project which aims to bring together quality improvement organisations from all four UK nations, however I was certainly struck by the potential of these QI experts working together as an alliance to share best (and worst!) practice across the whole spectrum of providers. Particularly interesting were the round table discussions which took place after lunch, where a different element of the improvement process was discussed at each table. One of these was hosted by the good old Quality Observa-tory, where Kate led conversations and encouraged ideas about the use of data and measurement to underpin improvement strategies. Although the room was filled with some fairly senior leaders from the world of healthcare, it was apparent that the contribution by us analytical types is vital to the improvement process.”

-Nikki Tizzard, Specialist Analyst

“I was very grateful to be able to attend the UKIA event as it was lovely to see that there were so many like-

minded people when it comes to improving quality within the NHS. I felt a bit out of place at first with the amount of people who had much more experience and expertise than myself however, as the morning went on I felt very

welcome and involved. The event was a great way to see the impact that our team can have (and does have) with other organisations from different areas of expertise on a wider scale rather than just what I've looked at and

been told. It's safe to say that this is not the last we have heard on improving quality, and I'm very proud and ex-cited to be part of it all!”

-Becki Ehren, Team Assistant

The round table discussions hosted by QO brought out some fascinating insights into the barriers to good measurement and provided some great ideas for how the UKIA could mobilise to assist. The main themes of dis-cussions included:

• Feeding measurement back into frontline teams

• Asking the right questions!

• Decoupling ‘data’ from ‘measurement’

• Measurement for improvement vs Measurement for judgement

• Measurement and analytical capacity

• Importance of capturing qualitative data and incorporating this along-side the quantitative piece

11


All for the good of the team6.

In September the QO team held a ‘team-building’ day, firstly to officially welcome our new recruits and secondly to remind ourselves what each other looks like, since we haven’t been able to work together in the same office since March 2013!

As it was a Friday we naturally kicked off proceedings with lunch in our local, then headed off to

nearby Crawley for a spot of bowling - boys vs girls! Now... technically the girls won, however the boys cried foul as they considered it cheating that some of the girls had used the ‘bumpers’ (which prevent the ball falling into the gutter).

Anyway, we eventually got bored arguing about it and went to the pub instead. We finished off the evening with a very fine meal, where we bonded over our shared love of honeycomb smash cheesecake!

The NHS Five Year Forward View

Simon Stevens – NHS Chief Executive – sets out his five year view for the NHS

The NHS Five Year Forward View was published on 23 October 2014 and sets out a vision for the future of the NHS.

This vision looks at the need for change and how best that this can be realised in the current socio-economic climate.

Further information is available from www.england.nhs.uk/ourwork/futurenhs/5yfv-

fore/

Potential Job Opportunities in the QO

The Quality Observatory team may be looking to expand further in the future, so if you’re interested in joining the best analytical team in the known universe (if we say so ourselves) please keep an eye out on the NHS Jobs website:

(https://www.jobs.nhs.uk/)

Kate Takes Great Strides for the Samaritans

On 21st September Kate ran 13.1 miles in an effort to raise £250 for The Samaritans’ work in East Surrey, in the first ever adidas Run Reigate Half Marathon.

So how did she get on? Six months of training paid off for Kate, who was one of 5,000 people running for charity. She crossed the finish line after a very respectable 2 hours and 32 minutes.

Well done Kate!

NEWS

Knowledge matters is the newsletter of NHS South East Coast’s Quality Observatory, to discuss any items raised in this publication, for further information or to be added to our distribution list, please contact us.

Hosted by: Central Southern Commissioning Support Unit

E-mail: [email protected]

To contact a team member: [email protected]

CELEBRATIONS

Birthdays Dani and Nikki have both had birthdays since the last issue and, as these pictures show, both received a bottle of something nice from the rest of the team.

Fascinating Facts

Surgeon Harold Gillies, horrified by the appalling injuries which resulted from shrapnel in World War 1, pioneered early techniques of facial reconstruction, and sparked the invention of plastic surgery.

Somewhere under my stomach, I'm quite thin

But you'd never know because I've got too much skin

Somewhere under my stomach, I'm petite,

It's just that all the ale and cakes are really indiscrete

Obesity's an epidemic

And the symptoms build and creep with every bite I eat

Where pounds pile on like building blocks

Away above the muffin tops, I'll have just one more sweet!

Somewhere under my stomach, I'm quite spry

Other people are quite healthy, why then, oh, why can't I?

If everyone can be fit and healthy

Why, oh why can't I?

Simon says>>.

During WW1 a US Army doctor, Captain Oswald Robertson, established the first blood bank in 1917. Sodium citrate was used to prevent the blood from coagulating and it was kept on ice for up to 28 days before being used in life-saving surgery.

Song for Obesity

(with apologies to Judy Garland>)

Congratulations to the Cheemas! Two of the pillars of the QO, Kate and Kiran Cheema, celebrated their 10 year wedding anniversary with a trip to Budapest, where it all began when they got engaged. Altogether now, aaaahh6...

Knowledge Matters Volume 8 Issue 4

Documents