Antimicrobial stewardship in care homes for older people development of a core outcome set

Antimicrobial stewardship in care homes for older people: development of a core outcome set by Hoa Quoc Nguyen, BPharm A thesis submitted to: Faculty of Medicine, Health and Life Sci

Antimicrobial stewardship in care homes for older people:

development of a core outcome set

by

Hoa Quoc Nguyen, BPharm

A thesis submitted to:

Faculty of Medicine, Health and Life Sciences

This thesis is dedicated to my late beloved grandfather, Doan Huu Tu,

and my late beloved supporter, Professor Liam Murray

4 The composition of the thesis is my own work

Table of contents

List of tables i

List of figures iii

Table of contents for appendices iv

Acknowledgments viii

List of abbreviations ix

Abstract xi

Publications xiii

CHAPTER 1 : General introduction 1

1.1 The ageing population 2

1.2 Older people and care homes 5

1.3 The need for antimicrobial stewardship in care homes 12

1.4 Core outcome sets 19

1.5 Overview of the research presented in this thesis 23

CHAPTER 2 : Interventions to improve antimicrobial stewardship for older people in care homes: a systematic review 25

2.1 Introduction 26

2.2 Aims and Objectives 28

2.3 Research design and methodology 29

2.4 Results 34

2.5 Discussion 55

2.6 Conclusion 62

CHAPTER 3 : Antimicrobial stewardship for older people in care homes: Outcomes of importance to researchers, healthcare professionals, and residents’ family members 63

3.1 Introduction 64

3.2 Aim and Objectives 65

3.3 Research design and methodology 66

3.4 Results 78

3.5 Discussion 98

3.6 Conclusion 106

CHAPTER 4 : Development of a core outcome set for clinical trials aimed at improving antimicrobial stewardship in care homes 107

4.1 Introduction 108

4.2 Aim and Objectives 109

4.3 Research design and methodology 110

4.4 Results 123

4.5 Discussion 138

4.6 Conclusion 148

CHAPTER 5 : Selection of outcome measurement instruments for a core outcome set for clinical trials aimed at improving antimicrobial stewardship in care homes 149

5.1 Introduction 150

5.2 Aim and Objectives 152

5.3 Research design and methodology 152

5.4 Results 166

5.5 Discussion 185

5.6 Conclusion 196

CHAPTER 6 : General discussion and Conclusions 197

6.1 General discussion 198

6.2 Conclusions 213

REFERENCES 214

APPENDICES 237

i

List of tables

Table 1.1 Common infections in care homes (Smith et al., 2008; Montoya and

Mody, 2011) 8

Table 1.2 Minimum requirements for an AMS programme (Fishman, 2012) 17

Table 2.1 Characteristics of included studies and effect of interventions 38

Table 2.2 Outcomes presented in included studies 47

Table 3.1 Overarching outcomes identified in a systematic review 78

Table 3.2 Participant demographics 79

Table 3.3 Outcomes deemed important by participants with illustrative quotes 81

Table 3.4 The refined list of outcomes related to AMS in care homes 96

Table 4.1 Inventory of outcomes by categories 124

Table 4.2 Demographic profile of participants in the Delphi panel 125

Table 4.3 Distribution of importance based on the scale used for each outcome in the first round 127

Table 4.4 Distribution of importance based on the scale used for each outcome in the second round 128

Table 4.5 Distribution of importance based on the scale used for each outcome in the third round 130

Table 4.6 Outcomes of importance and illustrative quotes after the first questionnaire 133

Table 4.7 Distribution of importance based on the scale used for outcomes after the online consensus exercise 135

Table 4.8 The core outcome set for use in trials aimed at improving AMS in care homes 137

Table 5.1 Measurement properties of an outcome measurement instrument (Prinsen et al., 2016a) 155

Table 5.2 Feasibility aspects defined by the COSMIN initiative (Prinsen et al., 2016a) 156

Table 5.3 Methodological quality of studies aimed at developing OMIs relevant to antimicrobial prescribing in care homes 170

ii

Table 5.4 Summary of quality assessment of ‘objective’ outcome measurement

instruments 172

Table 5.5 Summary of quality assessment of ‘subjective’ outcome measurement instruments 173

Table 5.6 Demographic profile of participants in the Delphi panel 177

Table 5.7 Distribution of agreement levels for each OMI in the first round 179

Table 5.8 Distribution of agreement levels for each OMI in the second round 180

Table 5.9 The COS developed for trials aimed at improving AMS in care homes and recommended OMIs 182

iii

List of figures

Figure 1.1 Percentage population change by age group across the UK, mid-2018 to

mid-2043 (Adapted from Northern Ireland Statistics and Research Agency, 2019) 3

Figure 1.2 Change in percentage of people prescribed at least one antibiotic by the time of admission to care homes (Taken from Patterson et al., 2019) 14

Figure 1.3 Process of development of a COS 21

Figure 2.1 PRISMA flow diagram of screening process and reasons for exclusion of studies published until November 30th 2018 35

Figure 2.2 PRISMA flow diagram of screening process and reasons for exclusion of studies published until December 31th 2019 from updated search 36

Figure 2.3 Risk of bias: review authors’ judgements about each risk of bias item presented as percentages across all included studies 54

Figure 2.4 Risk of bias summary: review authors’ judgements about each risk of bias item for each included study 55

Figure 3.1 The process of compiling a refined list of outcomes for AMS in care homes 97

Figure 4.1 Online consensus exercise overview 122

Figure 4.2 Three-round Delphi flow chart 131

Figure 5.1 Quality assessment of outcome measurement instruments 158

Figure 5.2 Process of screening and selecting articles to extract relevant OMIs 168

Figure 5.3 Two-round Delphi flow chart 180

iv

Table of contents for appendices

Appendix 2.1 PRISMA Checklist 238

Appendix 2.2 Search strategies 241

Appendix 2.3 Data extraction form 261

Appendix 2.4 ‘Risk of bias’ assessment of included studies 268

Appendix 3.1 Invitation letter for researchers and healthcare professionals 273

Appendix 3.2 Researchers and healthcare professionals’ information sheet 274

Appendix 3.3 Invitation letter for family members of care home residents 278

Appendix 3.4 Care home family members’ information sheet 280

Appendix 3.5 Topic guides for interviews with researchers and healthcare professionals 286

Appendix 3.6 Topic guide for family members’ interview 292

Appendix 3.7 Researchers and healthcare professionals’ consent form 298

Appendix 3.8 Family members’ consent form 299

Appendix 3.9 Demographic details form – Researchers and healthcare professionals 300

Appendix 3.10 Demographic details form – Family members of care home residents 301

Appendix 3.11 Letter of ethical approval 302

Appendix 3.12 Completed COREQ checklist 306

Appendix 3.13 Outcomes excluded from the refined list and reasons for exclusion 310

Appendix 4.1 Experts’ Round 1 questionnaire 313

Appendix 4.2 Public participants’ Round 1 questionnaire 318

Appendix 4.3 Template recruitment email 323

Appendix 4.4 Experts’ study invitation letter 324

Appendix 4.5 Experts’ information sheet (except for care home managers/ staff) 325 Appendix 4.6 Care home managers/ staff’ information sheet 330

Appendix 4.7 Public participants’ study invitation letter 335

Appendix 4.8 Public participants’ information sheet 337

v

Appendix 4.9 COMET Initiative Plain Language Summary 343

Appendix 4.10 Expert consent form for Delphi consensus (except for care home managers/ staff) 345

Appendix 4.11 Care home managers/ staff consent form for Delphi consensus 346

Appendix 4.12 Public participant consent form for Delphi consensus 347

Appendix 4.13 First round email for experts 348

Appendix 4.14 First round email for public participants 349

Appendix 4.15 Background Information and Completion Instructions for experts 350 Appendix 4.16 Background Information and Completion Instructions for public participants 352

Appendix 4.17 COMET Initiative Delphi Process Plain Language Summary 354

Appendix 4.18 Email reminder for experts 356

Appendix 4.19 Email reminder for public participants 357

Appendix 4.20 Second round email for experts 358

Appendix 4.21 Second round email for public participants 359

Appendix 4.22 Third round email for experts 360

Appendix 4.23 Third round email for public participants 362

Appendix 4.24 Experts’ invitation email for consensus group meeting 364

Appendix 4.25 Public participants’ invitation email for consensus group meeting 365 Appendix 4.26 New experts’ invitation for consensus group meeting 366

Appendix 4.27 New public participants’ invitation for consensus group meeting 367

Appendix 4.28 Experts’ information sheet for consensus group meeting 369

Appendix 4.29 Public participants’ information sheet for consensus group meeting 374

Appendix 4.30 Experts’ instructions of the online consensus exercise 379

Appendix 4.31 Public participants’ instructions of the online consensus exercise 381

Appendix 4.32 List of included outcomes after the Delphi consensus 383

Appendix 4.33 Experts’ consent form for consensus meeting 385

Appendix 4.34 Public participants’ consent form for online consensus exercise 386

Appendix 4.35 First questionnaire for the online consensus exercise 387

Appendix 4.36 Second questionnaire for the online consensus exercise 392

vi

Appendix 4.37 Letters of ethical approval 396

Appendix 4.38 Outcomes presented in the second-round questionnaire 398

Appendix 4.39 Response feedback of the first online questionnaire 399

Appendix 5.1 Search strategies 408

Appendix 5.2 COSMIN Risk of Bias checklist for studies developing outcome measurement instruments 411

Appendix 5.3 Criteria for good measurement properties 448

Appendix 5.4 Quality of evidence 450

Appendix 5.5 Researchers and healthcare staff (including care home staff) - Round 1 questionnaire 451

Appendix 5.6 Care home residents’ representatives - Round 1 questionnaire 457

Appendix 5.7 Invitation for researchers and healthcare staff (including care home staff) for online consensus exercise 462

Appendix 5.8 Invitation for care home residents’ representatives for online consensus exercise 463

Appendix 5.9 Information sheet for researchers and healthcare staff for online consensus exercise (except care home staff) 465

Appendix 5.10 Information sheet for care home staff for online consensus exercise 470

Appendix 5.11 Information sheet for care home residents’ representatives for online consensus exercise 475

Appendix 5.12 First round email for researchers and healthcare staff (including care home staff) 481

Appendix 5.13 First round email for care home residents’ representatives 482

Appendix 5.14 Background Information and Completion Instructions for researchers and healthcare staff (including care home staff) 483

Appendix 5.15 Background Information and Completion Instructions for care home residents’ representatives 485

Appendix 5.16 Email reminder for researchers and healthcare staff (including care home staff) 487

Appendix 5.17 Email reminder for care home residents’ representatives 488

vii

Appendix 5.18 Second round email for researchers and healthcare staff (including care home staff) 489Appendix 5.19 Second round email for care home residents’ representatives 490Appendix 5.20 Letter of ethical approval 491Appendix 5.21 List of outcomes included in the Core Outcome Set and outcome measurement instruments from literature search 492Appendix 5.22 Summary of 55 included studies 498Appendix 5.23 List of outcome measurement instruments (OMI) for quality

assessment 503Appendix 5.24 Quality assessment of feasibility aspects of outcome measurement instruments for the Core Outcome Set 508Appendix 5.25 Comments of participants after the first round of the Delphi

consensus exercise and responses of research team 512

viii

Acknowledgments

First and foremost, I would like to express my sincere gratitude to my two supervisors, Professor Carmel Hughes and Professor Michael Tunney I could have not completed my thesis without their guidance, patience, encouragement, and invaluable support I would also like to send my special thanks to Dr Declan Bradley, who gave me precious motivation and advice throughout my research It has been my great honour and privilege to be mentored by them

My PhD would have not started without immeasurable support of two important people: Professor Liam Murray and Ms Heather Taylor Liam had connected me with my current principal supervisor and continued to support my family in Belfast until his last day Heather has helped me since my first step of a new journey to Queen’s I would also like to thank my sponsor, the Vietnam International Education Development - the Ministry of Education and Training, for funding my PhD Studentship

My thesis was made possible thanks to brilliant and generous researchers who greatly assisted me without any hesitation: Dr Paula Tighe, Dr Yingfen Hsia, Dr Joseph Mylotte, Dr Heather Barry, Dr Audrey Rankin, and Dr Jacqueline Sneddon

I am also immensely grateful to all collaborators, supporters and participants who contributed their valuable time to the studies presented in the thesis My sincere thanks go to Marion and Stephanie, who greatly helped me with tape recordings

I really appreciate my PCRG friends and colleagues for their hospitality and friendship, Deborah, Mairead, Maureen, Nadia, Sarah, Caoimhe, Ameerah, Tahani, Dima, Lucy, Colleen, Bara’a, and Rineke My incredibly special thanks go to helpful staff at School of Pharmacy, especially Dr Roisin O’Hare, Dr Sharon Haughey and

Ms Sinead McCullough who have made my time at the school become meaningful Last but not least, I would like to extremely thank my parents, my brothers and my wife who have supported me during the hard time I especially send my grateful love

to my wife, Nga Nguyen, who has joined me for a grand journey to Belfast and who has been patient enough to make our dream come true

ix

List of abbreviations

CDI Clostridioides difficile infection

Instruments

Methods

The research presented in this thesis followed current guidelines for development of a COS (COMET) and selection of outcome measurement instruments (OMIs) for the COS (COSMIN) A systematic review was conducted to evaluate the effectiveness of randomised controlled trials (RCTs) to improve AMS in care homes and to determine outcomes reported in these studies A qualitative study was undertaken to explore outcomes deemed important to key stakeholders comprising researchers, healthcare professionals, and family members of care home residents Subsequently, a series of consensus procedures with relevant stakeholders were carried out to reach consensus

on outcomes that should be included in the COS for use in trials aimed at improving AMS in care homes Finally, methodology suggested by the COSMIN guidelines was followed to select OMIs available in the literature for outcomes in the COS

Results

The systematic review identified six RCTs with 28 outcomes reported and found the interventions had limited effect on improvement of antimicrobial prescribing

xii

Interviews with 41 key stakeholders in the qualitative study revealed 49 outcomes that were deemed important to them, including 40 additional outcomes which had not been used in previous RCTs Outcomes identified in the systematic review and the qualitative study were reviewed and refined by the research team to produce an inventory of 14 outcomes This inventory was presented in an online Delphi survey with 82 participants from 17 different countries, followed by an online consensus exercise with 12 participants in Northern Ireland Consensus was reached to include five outcomes in the COS, comprising two main outcomes (‘The total number of antimicrobial courses prescribed’ and ‘Appropriateness of antimicrobial prescribing’) and three optional outcomes (‘Days of therapy per 1000 resident-days’, ‘Rate of antimicrobial resistance’, and ‘Mortality related to infection’) Following the COSMIN approach, 17 potential OMIs for the COS identified in literature searches were assessed for quality of measurement properties Three OMIs with the best quality of evidence were presented in a consensus exercise with 59 Delphi panel members from 16 different countries Ultimately, consensus was reached to select two OMIs for two outcomes ‘The total number of antimicrobial courses prescribed’ and ‘Days of therapy per 1000 resident-days’ in the COS No OMIs were selected for the other outcomes in the COS

Conclusion

The research described in this thesis developed a COS for use in trials aimed at improving AMS in care homes Future studies should use this COS to measure the effectiveness and safety of AMS interventions in care homes

Nguyen HQ, Bradley DT, Tunney MM, Hughes CM Antimicrobial stewardship in

care homes: outcomes of importance to stakeholders Journal of Hospital Infection

2020;104(4):582-591 doi: 10.1016/j.jhin.2019.12.024

Nguyen HQ, Tunney MM, Hughes CM Interventions to Improve Antimicrobial

Stewardship for Older People in Care Homes: A Systematic Review Drugs & Aging

2019;36(4):355–369 doi: 10.1007/s40266-019-00637-0

Abstracts

Nguyen HQ, Tunney MM, Hughes CM Antimicrobial stewardship outcomes for older people in care homes: perspectives of researchers, healthcare professionals, and

family members of care home residents International Journal of Pharmacy Practice

2020;28(S1):46 [Conference: Health Services Research & Pharmacy Practice Annual Conference: Poster presentation]

Nguyen HQ, Tunney MM, Hughes CM Interventions to Improve Antimicrobial Stewardship for Older People in Care Homes 2018 [Conference: International Pharmaceutical Federation Annual Conference: Poster presentation]

1

CHAPTER 1: General introduction

2

1.1 The ageing population

1.1.1 Overview

Older people are defined differently in different countries according to chronological

age, changes in functional abilities or change in social role (Randel et al., 1999) The

United Nations defines older people as those aged 60 or over (United Nations, Department of Economic and Social Affairs, Population Division, 2015) However, the age of 65 years is commonly used as a cut-off to define older age-groups in most developed countries; at this age, the majority of the population have retired from paid

work (Rodrigues et al., 2012; United Nations, Department of Economic and Social

Affairs, Population Division, 2015)

According to current literature, the population of older people in the world is growing significantly A substantial increase of 48% in the number of people aged 60 and over between 2000 and 2015 was reported, and it has been estimated to reach 1.4 billion in 2030 (United Nations, Department of Economic and Social Affairs, Population Division, 2015) Europe and Northern America have witnessed massive

growth in the numbers of older people (Rodrigues et al., 2012; United Nations,

Department of Economic and Social Affairs, Population Division, 2015) In the United Kingdom (UK), a recent report from the Office for National Statistics indicated that there had been a shift toward population ageing since 2005 and the percentage of older population (aged 65 years and over) would increase from 17.8%

in 2015 to 24.6% in 2045 (Office for National Statistics, 2017) The number of older people in Northern Ireland (NI) is estimated to reach a quarter of the whole

3

population by 2043, which is faster compared to the rest of the UK (Northern Ireland Statistics and Research Agency, 2019) The percentage population change in pension age (the majority aged 65 or over) between 2018 and 2043 across the UK is presented in Figure 1.1

Children (aged 0-15) Working age (majority aged 16-64) Pension age (majority aged 65 and over)

Figure 1.1 Percentage population change by age group across the UK, mid-2018 to

mid-2043 (Adapted from Northern Ireland Statistics and Research Agency, 2019)

According to a World Health Organization (WHO) report, one of the main reasons for the expansion of the older population is the increase in life expectancy globally which results from better primary and secondary healthcare (World Health Organization, 2015b) This significant change in population dynamics requires governments to improve healthcare systems to adapt to the increasing needs of older people and maintain their well-being (United Nations, Department of Economic and Social Affairs, Population Division, 2015)

4

1.1.2 Characteristics of older people

Older people are more likely to suffer from both physical and mental health issues compared to younger adults The ageing process involves the gradual accumulation

of biological damages at cellular and molecular levels, thereby increasing the risk of functional and/or cognitive deterioration amongst older people (Niccoli and Partridge, 2012; World Health Organization, 2015b) Furthermore, other factors may also affect the ageing process such as lifestyle and environmental exposure

(Franceschi et al., 2018) Physiological changes due to normal ageing have been

recognised including sensory degradation, decreased muscle strength, increased fat mass, and immunosenescence (i.e gradual deterioration of the immune system due to aging) (Jaul and Barron, 2017) It has also been reported that diseases such as cardiovascular diseases, neoplasms, neurological and musculoskeletal disorders are the leading causes of disability and death amongst people aged 70 years and older,

(Tello et al., 2019) In addition, communicable diseases, such as lower respiratory

tract infections, are amongst the top five causes of lost years in populations aged 60 years and older in 2012 (World Health Organization, 2015b) Another common condition associated with disease burden amongst people at advanced age is the

‘geriatric syndrome’ which consists of falls, delirium, incontinence and frailty

(Inouye et al., 2007) Frailty is a state of vulnerability to sudden health status

changes triggered by a stressor event, which increases the risk of adverse outcomes

such as falls or disability (Clegg et al., 2013) The prevalence of frailty has been

estimated at approximately 5% amongst those aged 60 and over in the UK, and up to

65% in those aged 90 and over (Gale et al., 2015; Hanlon et al., 2018)

5

Due to the high risk of having comorbidities, the older population is also at increased risk of being prescribed multiple medications, or polypharmacy (World Health Organization, 2015b; Jaul and Barron, 2017) Polypharmacy may be appropriate but

it increases the likelihood of adverse drug events or drug interactions, which negatively impacts on quality of life of older patients (World Health Organization,

2015b; Jaul and Barron, 2017; Masnoon et al., 2017) Consequently, older people are

prone to a decline in their functional and cognitive capacity, which requires them to need substantial support from others

1.2 Older people and care homes

1.2.1 Care homes for older people

In the UK context, older people are provided home-based care by their carers, or facility-based care which involves care homes Care homes or long-term care facilities provide accommodation, services and personal support for older people, especially those who are very frail or dependent on other care in their everyday life Care homes include residential homes and nursing homes Nursing homes are distinct from residential homes as they also provide 24-nursing care for older people

who suffer from illness, injury or frailty (Statutory Instrument, 1992; Sanford et al.,

2015) Nursing homes can be divided into two subtypes: general nursing homes providing routine care and specialized nursing homes providing a specific type of

care (e.g care for residents with dementia) (Belan et al., 2020).In the United States

of America (USA), care homes are defined as long-term care facilities which are divided into skilled-nursing facilities and assisted-living facilities In Australia, care

6

homes are referred to as aged-care facilities providing either high-level or low-level

care (Alldred et al., 2016)

The number of care homes has been gradually increasing over time in response to the growth of the older population It has been estimated in the WHO European Region that the number of beds in care homes per 100,000 population increased from 383 in

1999 to 494 in 2013 (World Health Organization, 2016) In the UK, there are more than 11,000 care homes for older people from around 5,500 providers with available capacity of 454,000 beds (Competition & Markets Authority, 2017) However, the current number of care homes may not meet the demands of the older population who require long-term care A recent study has indicated a significant increase in the number of people aged 65 years and older with dependency, and estimated that an addition of 71,215 care home beds would be required by 2025 to meet the rate of

population ageing (Kingston et al., 2017)

1.2.2 Characteristics of older people in care homes

Older people living in care homes are very vulnerable The majority of care home residents suffer from a range of disabilities and comorbidities, resulting in a clinically complex population According to a recent systematic review, the mean prevalence of frailty in nursing homes was more than half of residents and the prevalence in some homes had been reported to approach 75% (Kojima, 2015) Older adults admitted to care homes are more likely to develop health conditions and

functional deficits, including high blood pressure and memory problems (Green et

al., 2017) In one study, approximately 60% of residents in residential homes

7

suffered from difficulties with mobility or immobility and up to 81% of residents had

comorbidities (e.g cardiovascular disease, diabetes, pulmonary disease) (McClean et

al., 2012) A study using data from general practices in England and Wales found

that the prevalence of stroke, dementia and mental illness in care homes was 2.5 to

15 times higher than that in the community (Shah et al., 2010) Another study on

health status of UK care home residents reported high proportions of dependency, cognitive impairment, multimorbidity, and polypharmacy in both residential and

nursing home settings (Gordon et al., 2014) In addition, care home residents are

more likely to have emergency admissions to hospital than those in the general population due to acute conditions such as pneumonitis, dementia, or pneumonia

(Smith et al., 2015) As the support needs of care home residents have increased over time (Green et al., 2017), it is essential to improve the quality of current healthcare

systems, including care homes

1.2.3 Infections amongst older people in care homes

Of particular note is the management of infection; care home residents are vulnerable

to infections due to physiological and cognitive changes, health status and routine

inter-person contacts among this population (Smith et al., 2008) Bacteria are the

most common cause of infections in care homes (Montoya and Mody, 2011) However, viral and fungal infections have also been identified amongst care home

residents (Falsey and Dallal, 2008; Montoya and Mody, 2011; Flevari et al., 2013; Lubeek et al., 2015) These residents may contract a variety of infections (Table 1.1)

but the three most prevalent infections reported in care homes are urinary tract

8

infections (UTIs), respiratory tract infections (RTIs), and skin and soft tissues

infections (SSTIs) (Smith et al., 2008; Montoya and Mody, 2011)

Table 1.1 Common infections in care homes (Smith et al., 2008; Montoya and

Mody, 2011)

Skin and soft tissues Pressure ulcers, cellulitis, scabies, herpes zoster and simplex

Gastrointestinal tract Viral, bacterial gastroenteritis, Clostridioides difficile infection

Urinary tract infections

UTIs comprise infections which occur anywhere along the urinary tract, such as cystitis, pyelonephritis, or prostatitis, and are very common in older adults (Mouton

et al., 2001; Detweiler et al., 2015) Typical signs and symptoms of UTIs include

dysuria, urinary frequency, suprapubic tenderness and fever but these may be absent

or vague in older patients (Mouton et al., 2001) Besides non-pharmacological

approaches, antibiotics are commonly prescribed empirically or based on the

detected pathogen to treat UTIs (Detweiler et al., 2015) However, asymptomatic

bacteriuria, the presence of bacteria in the urine without signs or symptoms of UTIs,

is not recommended to be treated with antibiotics amongst older people due to lack

of benefit and potential adverse effects (Nicolle et al., 2019)

9

UTIs are the most common infections and deemed the most over-diagnosed infection

in care homes (Smith et al., 2008; Montoya and Mody, 2011) Besides comorbidities

amongst care home residents, urinary catheterisation may contribute to the high prevalence of UTIs in this setting About 7% to 10% of care home residents have

indwelling urinary catheters (Smith et al., 2008) Longer duration of urinary catheter

insertion is associated with a higher risk of UTIs, and bacteriuria has been found in almost 100% of cases with urinary catheterisation after 30 days (Montoya and Mody, 2011) Older people with UTIs left untreated are at high risk of bloodstream infections and mortality, especially those with urinary catheterisation (Montoya and

Mody, 2011; Gharbi et al., 2019)

Respiratory tract infections

RTIs comprise infections which occur anywhere along the respiratory tract, including upper RTIs such as the common cold or rhinosinusitis, and lower RTIs such as bronchitis or pneumonia (Centre for Clinical Practice at NICE, 2008) Bacterial pneumonia and influenza are the most prevalent infections associated

with adverse outcomes amongst older people (Mouton et al., 2001; Montoya and

Mody, 2011) It has been estimated that people aged 65 years and over account for more than 80% of deaths due to influenza, and 60% of hospital admissions amongst

this population are due to pneumonia (Mouton et al., 2001) Antibiotics are

commonly used to treat pneumonia, whereas antivirals such as amantadine or

oseltamivir are recommended for treatment of influenza (Mouton et al., 2001)

Vaccination can reduce the risk of pneumonia and influenza in older population

(Mouton et al., 2001; Smith et al., 2008)

10

The prevalence of RTIs amongst older residents in care homes reported in the

literature is high, with levels reported up to 56% (Childs et al., 2019) Pneumonia

and other lower RTIs are largely responsible for hospitalisation and a leading cause

of mortality amongst care home residents (Montoya and Mody, 2011) It has been reported that care home-acquired pneumonia has a significantly higher mortality rate

than community-acquired pneumonia in the older population (Smith et al., 2008)

Moreover, outbreaks of influenza in care homes are frequently reported and often

severe (Smith et al., 2008; Montoya and Mody, 2011) The common occurrence of

RTIs in care homes could be explained by impaired immunity and comorbidities

amongst older residents as well as close contacts (Smith et al., 2008)

Skin and soft tissues infections

SSTIs are infections of skin and its supporting structures, including subcutaneous fat,

fascial layers, and musculotendinous tissues (Ramakrishnan et al., 2015) SSTIs in

the older population can vary from viral infections such as herpes zoster, to bacterial

infections such as cellulitis due to Staphylococci (Mouton et al., 2001; Anderson and

Kaye, 2007) Skin changes and immunosenescence due to ageing, comorbidities such

as diabetes, and a high frequency of conditions with skin fragility (e.g oedema or trauma) may explain why SSTIs are more common amongst older people (Anderson and Kaye, 2007) Antibiotics or antivirals are usually prescribed for treatment

depending on the cause of SSTIs (Mouton et al., 2001)

SSTIs are prevalent in care homes, especially infection of pressure ulcers Up to 20%

of care home residents have pressure ulcers which may subsequently result in SSTIs

11

(Smith et al., 2008; Montoya and Mody, 2011) Frail older people with pressure

ulcers are more likely to develop severe infections such as osteomyelitis and bacteremia (Montoya and Mody, 2011) Risk factors predisposing care home residents to pressure ulcers include immobility, focal pressures and shearing forces,

incontinence, and malnutrition (Smith et al., 2008) In addition, scabies caused by

mites is a common contagious skin infection reported in care homes, and this condition may lead to a secondary bacterial infection (Montoya and Mody, 2011)

Other infections

A number of diarrhoeal outbreaks in care homes have been reported due to viral (e.g

norovirus) and bacterial (e.g Escherichia coli) gastroenteritis (Smith et al., 2008; Montoya and Mody, 2011) Diarrhoea associated with Clostridioides difficile and asymptomatic colonisation of C difficile is also common in care homes (Montoya

and Mody, 2011) In addition, older people in care homes are also at high risk of contracting other infections such as conjunctivitis, bacteraemia, and endocarditis

(Smith et al., 2008)

A care home is an ideal reservoir of infections as food, services and environment are

shared by susceptible residents within a crowded setting (Strausbaugh et al., 2003)

Additionally, staff turnover and regular family visits may spread pathogens from

community and hospitals into this setting (Strausbaugh et al., 2003; Smith et al.,

2008) The two most common infections in long-term care, RTIs and UTIs, are associated with high rates of hospitalisation and mortality, prolonged hospital stay,

12

and substantial healthcare expenses (Juthani-Mehta and Quagliarello, 2010; Moro et

al., 2010; Montoya and Mody, 2011)

1.3 The need for antimicrobial stewardship in care homes

1.3.1 Antimicrobial prescribing in care homes

Antimicrobials which include antibiotics, antivirals and antifungals, are routinely prescribed to manage and treat infection in care homes The prevalence of antimicrobial prescribing in long-term care facilities varies across countries

(McClean et al., 2011; European Centre for Disease Prevention and Control, 2014)

According to the European Centre for Disease Prevention and Control (ECDC), the mean prevalence of antimicrobial use in European long-term care facilities in 2013 was 4.4%, but prevalence was more than 10% in some countries, such as the Czech Republic, Denmark and NI (European Centre for Disease Prevention and Control, 2014) This number increased to 4.9% in the latest point prevalence study of antimicrobial use amongst European care homes in 23 countries in 2016 and 2017, and six countries, also including NI, had a mean prevalence of antimicrobial use

greater than 10% (Ricchizzi et al., 2018)

The proportion of care home residents exposed to at least one antimicrobial course

annually or during study periods were high (50% to 80%) (Benoit et al., 2008; Lim et

al., 2014; Gillespie et al., 2015) A study exploring antimicrobial use across care

homes in the UK found that UTIs, RTIs, and SSTIs were the three most common

indications for antimicrobial prescriptions (Thornley et al., 2019) Moreover, up to 47.1% of prescriptions for UTIs were for prophylactic purposes (Thornley et al.,

13

2019) Broad-spectrum antimicrobials were frequently prescribed for older people in care homes The ECDC reported that beta-lactams were most commonly used to treat infections, including penicillins with extended spectrum and beta-lactamase resistant penicillins; quinolones, sulfonamides and trimethoprim were also frequently

prescribed (European Centre for Disease Prevention and Control, 2014; Ricchizzi et

al., 2018) Similarly, beta-lactams, quinolones, and sulfonamides and trimethoprim

were the most frequently prescribed antimicrobials in long-term care facilities

according to a number of studies in a review (Lim et al., 2014)

Furthermore, older people in care homes are more likely to receive antimicrobial prescriptions than those living in community A study in NI found that care home residents were twice as likely (adjusted odds ratio: 2.05) to be prescribed an

antibiotic compared to community dwellers (Patterson et al., 2019) Figure 1.2

presents the change in percentage of people receiving an antibiotic before and after

admission to care homes reported by Patterson et al during January 2012 to December 2013 (Patterson et al., 2019) Another study reported that residents in care

homes were more likely to receive higher numbers of antibiotic prescriptions for

UTIs than those not in care homes (Sundvall et al., 2015) In addition, high

prevalence of antimicrobial use in this setting was associated with care homes having fewer than 65 beds, residents aged over 85 years, male gender, residents with urinary

or vascular catheters, and residents with surgery in the previous 30 days (Ricchizzi et

al., 2018).

14

Figure 1.2 Change in percentage of people prescribed at least one antibiotic by the

time of admission to care homes (Taken from Patterson et al., 2019)

High rates of inappropriate prescribing in terms of dosage, treatment duration, decision to initiate or withhold antimicrobials, and regimens have been reported in

care home settings McClean et al found that up to 25% of antimicrobials were

prescribed at inappropriate doses (i.e higher than recommended doses) in care homes, and 25% to 55% of all antimicrobials were initiated by telephone

consultations in residential homes (McClean et al., 2011; McClean et al., 2012) Telephone antimicrobial prescribing is also common in nursing homes; Schweizer et

al found that most general practitioners admitted that they seldom visited care home

patients and they accepted the assessment undertaken by nursing staff which often

suggested the need for antimicrobials (Schweizer et al., 2005) A study in the

Netherlands indicated that a quarter of overall antibiotic treatment decisions were

15

inappropriate, especially with respect to treatment of UTIs (van Buul, Veenhuizen, et

al., 2015) Additionally, Peron et al reported that up to 49% of care home residents

received at least one day of unnecessary antimicrobial treatment (Peron et al., 2013) Eure et al also reported that less than 50% of nursing home residents were given appropriate antibiotic initiation and administration for UTIs (Eure et al., 2017)

Inappropriate antimicrobial prescribing may potentially lead to serious adverse drug reactions or dangerous interactions due to physiological changes in organ

systems and polypharmacy in older people (Corsonello et al., 2015) Antimicrobial consumption has also been associated with Clostridioides difficile infection (CDI)

(Fletcher and Cinalli, 2007), and care home residents are at a higher risk of CDI because of advanced age and receiving long-term healthcare (Chopra and Goldstein, 2015) Moreover, frequent antibiotic use, especially broad-spectrum antibiotics, has been linked to the development of resistant pathogens (World Health Organization, 2015a)

1.3.2 Antimicrobial resistance in care homes

Antimicrobial resistance (AMR) is defined as the state that microorganisms, such

as viruses, bacteria, and fungi, are no longer susceptible to antimicrobials used to treat them due to developed or acquired resistance genes (National Institute for

Health and Care Excellence, 2015b; World Health Organization, 2015a; Holmes et

al., 2016) This phenomenon is a natural evolutionary process under the pressure of

antimicrobial exposure but it occurs quickly with antimicrobial use in healthcare

and agriculture (Holmes et al., 2016) A major concern with development of AMR

16

is that there are fewer or no options available to treat infections caused by resistant organisms (World Health Organization, 2014) Consequently, antimicrobial resistant infections result in increased adverse outcomes in patients and increased treatment costs (World Health Organization, 2014; Center for Disease Dynamics, 2015) The strong association between antimicrobial prescribing and AMR development has been established with frequent antimicrobial use increasing the

likelihood of AMR development (Schechner et al., 2013; Bell et al., 2014)

Care homes have been recognised as reservoirs of antimicrobial resistant organisms

such as methicillin-resistant Staphylococcus aureus (MRSA), multidrug resistant gram-negative bacteria, and vancomycin-resistant Enterococci (Esposito et al., 2007; van Buul et al., 2012) AMR may be introduced into a care home in two different

ways People outside the care home settings who have been colonised or infected by resistant organisms may spread these pathogens to care home staff and residents; or the resistant organisms may develop in care home residents who have received

multiple courses of antimicrobials (Nicolle et al., 1996) Antimicrobial resistant

pathogens are deemed to complicate treatment, reduce quality of life and increase mortality worldwide (World Health Organization, 2014); residents in care homes

may suffer from the same poor outcomes (van Buul et al., 2012), albeit with more serious consequences because of their susceptibility to infection (Smith et al., 2008)

A study in NI showed that MRSA was detected among both residents and healthcare staff in nursing homes Residents were more likely to be colonized if they lived in homes in which more than 12.5% of screened staff were colonized with MRSA

(Baldwin et al., 2009) AMR in care homes may lead to substantial treatment

17

expenses, impact on quality of life, increased risks of hospitalisation and mortality

(Esposito et al., 2007; van Buul et al., 2012; Fleming et al., 2016)

1.3.3 Antimicrobial stewardship in care homes

Antimicrobial stewardship (AMS) is defined as an elaborate intervention to improve the appropriateness of antimicrobial prescribing so as to optimise treatment, reduce adverse outcomes and alleviate AMR (Fishman, 2012; National Institute for Health and Care Excellence, 2015b) Minimum requirements recommended for an AMS programme are presented in Table 1.2

Table 1.2 Minimum requirements for an AMS programme (Fishman, 2012)

Creation of an AMS

team

interprofessional, including (but not limited to) a physician, a pharmacist, a clinical microbiologist, and an infection preventionist At least one member of the team should have training in AMS

demonstrated clinical need should be available

management of common infections

susceptibilities to key pathogens periodically

18

AMS strategies, along with clinical guidelines, have been implemented in hospitals, but there have been calls to extend this to primary care settings, including care homes (Rhee and Stone, 2014) In the latest guidelines, the National Institute for Health and Care Excellence (NICE) identified care home staff as a target group to implement AMS (National Institute for Health and Care Excellence, 2015b) The Infectious Diseases Society of America (IDSA) has also recommended the implementation of

antibiotic stewardship programmes in nursing homes (Barlam et al., 2016)

Considering the negative impact of antimicrobial resistant infection and issues with antimicrobial prescribing on older people in care homes, it is essential to introduce AMS in this setting (Rhee and Stone, 2014) Recently, the Agency for Healthcare Research and Quality (AHRQ) and the Centers for Disease Control and Prevention (CDC) have published guidelines and tools for nursing homes to implement AMS (Centers for Disease Control and Prevention, 2015; Agency for Healthcare Research and Quality, 2016) In a recent technical report, the ECDC has also proposed the establishment of AMS activities with multifaceted approaches in care homes to be included in future guidelines on antimicrobial use in humans (European Centre for Disease Prevention and Control, 2017) However, it is noted that there are challenges

to such interventions in care homes, including limited access to healthcare specialists and modern diagnostic equipment, ambiguous presentation of infection in older

people, and lack of infection prevention and control strategies (Lim et al., 2014; Morrill et al., 2016) There are also challenges in assessing the overall effectiveness

of interventions due to heterogeneity in outcomes as discussed in Section 1.4

19

1.4 Core outcome sets

1.4.1 Rationale for core outcome sets

There has been concern about the quality and quantity of measured outcomes in trials Inconsistencies in reported outcomes has impeded systematic reviewers in collating and combining findings (Clarke and Williamson, 2016) Another study found substantial heterogeneity in 372 outcomes reported in 47 trials of medication reviews in older patients, and also indicated insufficient evaluation of some important outcomes such as adverse events and patient-reported outcomes across the

studies (Beuscart et al., 2017)

The Core Outcome Measures in Effectiveness Trials (COMET) initiative was established in 2010 to facilitate development and application of core outcome sets

for clinical trials (Williamson et al., 2017) A core outcome set (COS) is defined as

a standardised list of outcomes achieved by consensus that represents the minimum

to be reported in all trials of a specific health area (Williamson et al., 2012) It is

hoped that a COS developed for a particular area should be used in all future trials, thereby facilitating the comparability of the effectiveness across these trials, and improving the quality of evidence from a systematic review or meta-analysis

(Williamson et al., 2017)

1.4.2 Development of core outcome sets

After establishing the need for a COS, the main process to develop the COS consists

of compiling an inventory of potential outcomes through existing knowledge and

20

literature, and conducting a consensus exercise involving relevant stakeholders to

produce a final COS (Williamson et al., 2017) In order to identify important

outcomes, which are meaningful not only to researchers but also patients and carers,

a number of approaches are recommended including undertaking systematic reviews

and qualitative studies involving relevant stakeholders (Williamson et al., 2017)

While systematic reviews identify reported outcomes which may be of interest to researchers and perhaps policy makers, qualitative approaches explore views of other stakeholders, such as healthcare professionals and patients, regarding outcomes that are important to them In addition, qualitative approaches seek to understand why identified outcomes are important, and to identify appropriate language for labelling and describing outcomes in a subsequent consensus exercise of COS development

(Keeley et al., 2016)

For instance, a study developing a COS for trials to optimise prescribing in care homes identified a long list of 63 outcomes from a published literature review and stakeholder input, and achieved a set of thirteen prioritised outcomes through a two-round online Delphi exercise in which a wide range of stakeholders

participated (Millar et al., 2017) Another exemplar is a COS for reconstructive

breast surgery which was developed by compiling a list of outcomes through systematic reviews and stakeholder interviews, implementing a two-round Delphi consensus, and holding a face-to-face meeting to generate a set of eleven key

outcomes (Potter et al., 2015)

After a COS in a specific area has been developed (what to measure), it is also essential to identify which measurement instruments should be used to measure

21

outcomes in the COS (how to measure) (Williamson et al., 2017) The

Consensus-based Standards for the selection of health Measurement Instruments (COSMIN) initiative, founded in 2005, has facilitated development and quality assessment of

outcome measurement instruments (OMIs) (Mokkink et al., 2016) An OMI is

defined as a tool which measures a specific outcome regarding quality or quantity

(Prinsen et al., 2016b) For instance, an outcome ‘pain intensity’ can be measured by

using one of three OMIs: Visual Analogue Scale VAS, Numeric Rating Scale NRS,

or Pain Severity subscale of Brief Pain Inventory BPI-PS (Chiarotto et al., 2018) In

2016, the COMET and the COSMIN initiatives published joint guidelines

recommending how to select OMIs for outcomes in a COS (Prinsen et al., 2016b)

The process for development of a COS is summarised in Figure 1.3

Figure 1.3 Process of development of a COS

et al., 2016) Similarly, a systematic review to determine components of a successful

intervention from RCTs on antibiotic prescribing in long-term care facilities found that interventions involving local guideline development or educational strategies improved antibiotic prescribing modestly due to poor quality of evidence and mixed

results (Fleming et al., 2013)

From the current literature, there have been a number of outcomes that were

measured and reported in studies of AMS for older people in care homes (Fleming et

al., 2013; Morrill et al., 2016; Feldstein et al., 2018) Moreover, some outcomes such

as antimicrobial consumption measured by defined daily doses (DDDs) should be

considered carefully due to possible misinterpretation (Neilly et al., 2017) It is

essential to generate a COS to facilitate interpretation of findings and effectiveness

of interventions aimed at AMS in care homes