1

Assignment 3: Final Project - Comprehensive Literature Review

Jay Jalali

Yorkville University

PSYC6213: Research Methods

Dr. Ivana Djuraskovic

November 17th, 2020

 2

Abstract

The comprehensive literature review explores research from 6 quantitive sources and 6

qualitative studies conducted over the last ten years on the use of Cognitive Behavioural Therapy

(CBT) and Physical Exercise (PE) interventions in Pain Self Management Programs (PSM) for

the treatment of various types of chronic pain. The purpose of the review was an overall

evaluation of program efficacy, components that contribute to program experience and

challenges to long term commitment of participants to practice the intervention.

Keywords: chronic pain, self management, CBT

 3

The use of Cognitive Behavioural Therapy (CBT) and Exercise for Pain Self Management

(PSM) Programs

Chronic pain is the leading cause of disability and a public health issue prevalent in

almost 20% of adults globally (Rice, 2016). It is defined as pain experienced for three months or

longer and is classified into primary pain, in which pain is a primary presenting feature, and

secondary pain, which includes pain persisting after surgery, trauma or associated with an

identified disease or its treatment (Treede 2019).

Due to the complex nature of chronic pain, unimodal or pharmacological interventions

individually targeting the physical component of pain are often short term or non effective (Turk

et al., 2009). However, self pain management programs involving Cognitive Behavioural

Therapy (CBT) and Physical Exercise (PE) have been shown to be efficacious approaches to

symptom management as they address both psychological and functional components of

managing pain (William et al., 2020; Devan et al., 2018).

Background

CBT for PSM follows a rationale that patients must understand cognitive and behavioural

influences of pain in order to combat and self manage pain related emotions and behavior (Bruns

et al., 2006). By applying psychological principles to recognize maladaptive feelings, thoughts

and behaviours; CBT encourages patients with chronic pain to conceptualize pain as

manageable, gain an active role in controlling it and develop adaptive behavioural and cognitive

responses to functioning with pain (Bruns et al., 2006). Limited research additionally supports

this rationale, by combining CBT and interventions involving physical exercise (PE), to optimise

pain management and improve overall quality of life (Dysvik et al., 2004). While CBT targets
 4

psychosocial components, PE targets the biological components of chronic pain by improving

functionality and overcoming physical deconditioning (Dysvik et al., 2004).

Presently, not many studies have specifically examined the impact of combining specific

types of CBT and PE interventions on chronic pain and further research is required to evaluate

the intervention’s efficacy over long term periods (Natvig et al., 2010). This is because research

on PSM programs using both CBT and PE are sparse and comparisons are limited due to

differences in types of pain, service periods, program content, program intensity, result

evaluation approaches, and follow-up periods used (Joos et al., 2004). For the purpose of this

literature review individual and group PSM programs using qualitive and quantitate methods

involving adaptations of CBT and PE for various forms of pain, have been reviewed from the

last 10 years (William et al., 2020; Devan et al., 2018).

Problem Statement

With existing literature on adaptations of CBT and PE use in PSM programs varying in

relation to types of pain treated, target population, length of program and follow up period; the

purpose of this literature review is an overall evaluation of program efficacy, components that

contribute to program experience and challenges to long term commitment of participants to

practice the intervention.

Literature Review

Scope of Research

6 qualitative (Charest et al., 2015; Peters et al., 2016; Nordin et al., 2013; Dekker et al.,

2014; Barker et al., 2012; Furnes et al., 2014) and 6 quantitative studies (Pincus et al., 2015;

Rocca et al., 2017; Wood et al., 2013; Foti et al., 2016; Arakoski et al., 2015; Oey et al., 2016)

were conducted within hospitals, physical rehabilitation centres, secondary care clinics and
 5

research institutes in the United Kingdom, Italy, Canada, Australia, Finland, USA, Sweden and

the Netherlands.

Besides any individual inclusion and exclusion screening criteria, participants were

included based on type of pain experienced, have reported chronic pain for over 3 months

duration and referral from their medical practioner. The quantitive and qualitative studies

included participants that were aged 18 and over and 16 and over respectively. This inclusion of

studies based on a range of developed countries, treatment locations and vast participant age

groups allowed a broader scope of intervention review based on varying adaptations of treatment

options for pain type, length of intervention time, and the duration of post intervention follow up,

relative to the program.

In relation to types of pain, quantitive studies evaluated PSM programs for neck pain

(Rocca et al., 2017), low-back pain(Pincus et al., 2015; Foti et al., 2016), geriatric pain (Wood et

al., 2013), osteoarthritis (Arakoski et al., 2015) and general chronic pain (not limited to pain

type) (Oey et al., 2016); while qualitative studies evaluated fibromyalgia (Charest et al., 2015),

orofacial pain (Peters et al., 2016), low back pain (Barker et al., 2012) and general chronic pain

(Nordin et al., 2013; Dekker et al., 2014; Furnes et al., 2014). All studies were only selected if

the PSM program had a psychological based CBT treatment framework covering over 50 percent

of the program content, included some form of PE education and was delivered and supervised

by qualified psychological and physiotherapy field trained professionals.

Research Methods

5 quantitative studies selected used experimental research designs with randomised

control trial groups (Pincus et al., 2015; Rocca et al., 2017; Wood et al., 2013; Foti et al., 2016;

Arakoski et al., 2015) and 1used a quasi experimental design (Oey et al., 2016). The 5 qualitive
 6

studies used phenomenological (thematic) (Charest et al., 2015; Peters et al., 2016; Nordin et al.,

2013; Dekker et al., 2014; Furnes et al., 2014) and 1 grounded theory (Barker et al., 2012) study

design and analysis methods.

Quantitative Studies Review

The experimental method randomised control trial (RCT) studies revealed a range of

results in relation to outcomes of PSM programs on completion over a range of follow up periods

(Pincus et al., 2015; Rocca et al., 2017; Wood et al., 2013; Foti et al., 2016; Arakoski et al.,

2015). The advantages of reviewing these experimental studies with the RCT group as an

independent variable is that it allowed observations of any changes in participants where the

intervention was withheld (Goodwin & Goodwin, 2017). Since statistical significance test results

were readily interpretable in these studies, it increased the reliability of results by reducing the

likelihood of type one and type two errors (Goodwin & Goodwin, 2017). Additionally,

randomisation of participants also minimised any confounding factors because of unequal

distribution of prognostic factors, making different groups comparable to known and unknown

factors obtained during research (Goodwin & Goodwin, 2017).

The disadvantages of this method were that the long-term logistics and participation of

both participants and practioner of the intervention were significantly distorted due to lack of

consistent participation, dropping out of participants and changing the practioner between studies

(Pincus et al., 2015; Rocca et al., 2017; Wood et al., 2013; Foti et al., 2016; Arakoski et al.,

2015). Additionally, the ethical considerations were that some studies never considered a

thorough medical history of each participant in relation to previous pain treatment. There were

participants included that have had tried multiple prior pain treatments and were thus desperate

for a positive outcome. This made generalisation of results difficult because the samples seem to
 7

have people who enrol or get referred to studies seeking positive results (Pincus et al., 2015;

Rocca et al., 2017; Wood et al., 2013; Foti et al., 2016; Arakoski et al., 2015).

With the quasi experimental study with 36 different groups over 10 years, the pre-post

results showed significant increases in improved general health between completers and non

completers of programs (Oey et al., 2016). The advantage of this method is that the larger sample

base increased external validity of results and laid the foundation for future randomised studies

(Goodwin & Goodwin, 2017). The disadvantage to this method was the existence of non

completion bias from incomplete pre and post-test outcome measure questionnaires not being

checked by instructors for competedness (Oey et al., 2016). Additionally, the generalization of

findings was somewhat reduced reduced because of more female than male participants in

intervention groups (Oey et al., 2016).

Qualitative Results Review

The phenomenological studies included face to face group interviews and semi structured

interviews, with open ended questions for theme development of program elements (Charest et

al., 2015; Peters et al., 2016; Nordin et al., 2013; Dekker et al., 2014; Furnes et al., 2014). The

advantage of this method is that it did not seek to prove anything about the efficacy of PSM

interventions, instead it added to existing knowledge about interventions in certain populations

(Richards & Morse, 2013). Since these studies were thematic in purpose it improved the

confidence of results because it was based on participant’s own words and narratives than any

researcher bias with interpretations (Richards & Morse, 2013). The disadvantage with this

method is that all the studies used samples that contained no more than 34 participants. The low

sampling number although rich in internal validity to individual studies, were not highly

externally valid because of the non wider generalisation of findings (Richards & Morse, 2013).
 8

Additionally, an ethical consideration with these research studies was the lack of member

checking or allowing participants to clarify intentions with answers, correct errors or provide

additional information; before publishing final interpretations and conclusions (Charest et al.,

2015; Peters et al., 2016; Nordin et al., 2013; Dekker et al., 2014; Furnes et al., 2014).

The grounded theory study used semi structured interviews to comparative analysis of

pre, during and post program outcomes to identify narratives between a group of achieved

successful PSM results and a group that perceived no benefit (Barker et al., 2012).

The advantage of this method was increased credibility and richness of results because of the

triangulation of sources involving data obtained from comparative groups of participants at

different points in time (Richards & Morse, 2013). Although grounded theory offers theoretical

insight and not reproducibility, all the participants from the study were from one pain

management program and this reduced the transferability of the theory obtained to other pain

types and demographics (Barker et al., 2012).

Acceptance and Coping with Chronic Pain

All quantitive study patients for lower back, neck pain, geriatric and general pain

revealed results that the group-based intervention were higher in acceptance of pain, lower

distress, disability and fear avoidance; more than the standard intervention of physiotherapy and

general exercises alone used in the control groups (Pincus et al., 2015; Rocca et al., 2017; Wood

et al., 2013; Foti et al., 2016;Oey et al., 2016).While the osteoarthritis study revealed there was

no hypothesis proven of the intervention effectiveness over the ordinary GP care group, with

negative impact seen in self efficacy of pain in control groups and smaller decline in emotional

well being for those in the intervention programme (Arakoski et al., 2015).
 9

Among qualitive studies a hope and collective belief narrative was fundamental to groups

and was likely linked to rehabilitation and recovery. This was supported in content

interpretations that patients with no hope made no recovery and patients with a positive outlook

had restored hope with living with pain. This restoration of hope of patients hinged upon the

programs ability to deconstruct their fears, provide acceptable understanding of pain and

reconstruction of self ability (Charest et al., 2015; Peters et al., 2016; Nordin et al., 2013; Dekker

et al., 2014; Barker et al., 2012; Furnes et al., 2014).

A common theme among all types of studies were that they built new cognitive

behavioural skills and self acceptance skills including belief modification and challenging

maladaptive cognitions to induce positive attitudes toward recovery. Additionally, pain

perception had decreased in participants by the end of treatment, suggesting the importance of

functional exercises and cognitive behavioural therapy in sustaining modified pain perception

effectively in chronic populations.

Group Support and Shared Understanding of Chronic Pain

All qualitive studies revealed common themes that encounters with the group facilitators

and fellow participants were fundamental to acceptance and productive outcomes; from

increased empowerment, feeling identified, believed and obtaining relatable explanations on

treatments. A lot of participants had frequently undergone various consultations to investigate

underlying causes for their pain condition and previously felt stigmatised and not believed by

others (clinicians, family members, friends and acquaintances), when sharing their pain

experiences. Thus, feeling understood and believed during the program was essential to feeling

comfortable discussing and validating their symptoms and engage without feeling judged in the
 10

intervention (Charest et al., 2015; Peters et al., 2016; Nordin et al., 2013; Dekker et al., 2014;

Barker et al., 2012; Furnes et al., 2014).

In the long term strong therapeutic alliance was built over repeated sessions from

continual nonjudgmental and open exchanges, feeling valued, guided problem solving and

shared decision making. Additionally, a supportive ambience from clinicians, participants

family, friends, and work environment positively influenced “self-discovery” and made

individuals “feel empowered” to continue to use the strategies following the intervention outside

the program (Pincus et al., 2015; Rocca et al., 2017; Wood et al., 2013; Foti et al., 2016;

Arakoski et al., 2015; Oey et al., 2016).

Tertiary Topic Barriers to Self Pain Management

Some qualitative findings revealed that the day to day self management of chronic pain

post program completion has practical challenges in relation to time dedication needed to

practice CBT techniques and practicing exercises learnt, in isolation versus while in group

settings (Charest et al., 2015; Peters et al., 2016; Nordin et al., 2013; Dekker et al., 2014; Barker

et al., 2012; Furnes et al., 2014). Furthermore, avoiding or practicing the challenging thoughts of

pain caused mental conflict, which at times reduced coping effects from increased focus on pain

related thoughts. In some participants the lack of commitment with practicing the strategies post

program led to feelings of guilt and increased self criticism in addition to pain acceptance being

perceived as surrendering to pain or letting it control them.

It was understood that the meaning and intervention act of accepting pain seemed

counterproductive to those holding fixed biomedical beliefs towards their pain and this led to

continual inner conflict from focusing on pain reduction, leading to perceptions of the program
 11

as unhelpful (Charest et al., 2015; Peters et al., 2016; Nordin et al., 2013; Dekker et al., 2014;

Barker et al., 2012; Furnes et al., 2014).

Summary

The overall review of all studies revealed that the there were efficacious therapeutic

outcomes in participants experiencing different pain types, treated in various types of centres

during and after the use of adaptations of CBT and PE for PSM programs. This literature review

also demonstrated the various components that contribute to program experience and challenges

to long term commitment of participants to practice the intervention.

Implications

Since PSM programs incorporate fundamentally person-centered communication

approaches, which are a critical component of CBT centered care and shared decision making,

the review implicates the need for improved communication among stakeholders (i.e., patients,

clinicians, family, and friends) to increase the overall efficacy of self-management strategies

(William et al., 2020; Devan et al., 2018). For patients, interventions can incorporate strategies to

enhance communication skills; allowing participants more productive discussions and

comfortable engagement with instructors, to articulate personal experiences of pain, preferences

with treatment goals, and clarifications on their personal pain management plans (William et al.,

2020; Devan et al., 2018).

For psychologists, therapists, physiotherapists and practioner instructing in these

multidisciplinary interventions, the upskilling of communication skills to understand wider pain

populations should be introduced to allow for more shared understandings and validation during

sessions. Additionally, if instructors regularly acknowledge a participant’s efforts to manage pain

and recognize biopsychosocial struggle needed for continual coping, this acknowledgement and
 12

empathetic approach could possibly grow trust in the intervention and strengthen therapeutic

alliances by low drop out rates (William et al., 2020; Devan et al., 2018).

The main themes and learnings offered by these research outcomes from different

countries, chronic pain types and treatment centres should now add to the existing knowledge

needed to: (a) empower decisions of managing pain and maintaining sustainable lifestyles within

a more informed biopsychosocial model, (b) tailor treatments to pain types and individual needs,

and (c) synergise multidisciplinary expertise of physical therapists and psychologists in order to

effectively co-deliver these programs consistently (William et al., 2020; Devan et al., 2018).

More practically, this means research outcomes can be applied to make programs more daily

living applicable and transferable across pain types; by screening patients and matching their

specific requirements to interventions and continually seeking to develop and test cross-

disciplinary with shared delivery outcomes for patient progress (William et al., 2020; Devan et

al., 2018).

Ideas for Future Research

An idea for improvement of future quantitive research would be the use of standard

measures to improve comparability across studies (Goodwin & Goodwin, 2017). Although some

studies used commonly accepted measures for pain, catastrophizing and fear outcomes, there

were a range of different measurement methods for subjective experiences, theme variations

within outcomes, and behavioural measures supplementing any qualitative self-reporting.

A potential quantitative research bias problem worth noting here is many experimental

studies were designed in a way that primary outcomes were psychological abstraction

characterising presumed process (for example, acceptance of pain) rather than tangible outcomes

of living with pain (Pincus et al., 2015; Rocca et al., 2017; Wood et al., 2013; Foti et al., 2016;
 13

Arakoski et al., 2015; Oey et al., 2016). Oftentimes such inferential abstractions are

misunderstood by patients and variably measured by researchers, since chronic pain patients

identify with a varied range of outcomes compared to what is currently assessed (Biguet 2016).

With the quality of research reviewed and various trials of CBT and PE used in PSM

programs, there is less discovery value for any further randomised controlled trials (RCTs) and

phenomenological studies for standard CBT in the management of pain. Glasziou & Chalmers

(2018) suggest that resources to investigating the efficacy of CBT in people with chronic pain

should not be allocated to small trials, regardless of pain condition, since research waste is a risk.

Research resources should instead be targeted to larger multi-centre studies for different types of

psychological therapies for people with chronic pain, including extension of established CBT

methods in under-represented pain populations (Glasziou & Chalmers, 2018).

Additionally, all trials can allow individual patient data to be made available to facilitate

individual patient analysis and pooling of data to identify outcome variances, since some studies

may suggest ways to maximise treatment benefits that were previously untested(William et al.,

2020). This will be beneficial because there is extensive preclinical interest and research in

behavioural treatment for PSM but any advancement to into clinical level studies have been

small scale (William et al., 2020).

 14

References

B. Oosterhof, J. H. M. Dekker, M. Sloots, E. A. C. Bartels & J. Dekker (2014) Success or failure

of chronic pain rehabilitation: the importance of good interaction – a qualitative study

under patients and professionals, Disability and Rehabilitation, 36:22, 1903-1910.

https://doi.org/10.3109/09638288.2014.881566

Boschen, K. A., Robinson, E., Campbell, K. A., Muir, S., Oey, E., Janes, K., Fashler, S. R., &

Katz, J. (2016). Results from 10 Years of a CBT Pain Self-Management Outpatient

Program for Complex Chronic Conditions. Pain research & management, 2016,

4678083. https://doi.org/10.1155/2016/4678083

Bourgault, P., Lacasse, A., Marchand, S., Courtemanche-Harel, R., Charest, J., Gaumond, I.,

Barcellos de Souza, J., & Choinière, M. (2015). Multicomponent interdisciplinary group

intervention for self-management of fibromyalgia: a mixed methods randomized

controlled trial. PloS one, 10(5), e0126324. https://doi.org/10.1371/journal.pone.0126324

Bruns, D., & Disorbio, J. M. (2009). Assessment of Biopsychosocial Risk Factors for Medical

Treatment: A Collaborative Approach. Journal of Clinical Psychology in Medical

Settings, 16(2), 127–147. doi:10.1007/s10880-009-9148-9

Devan, H., Hale, L., Hempel, D., Saipe, B., & Perry, M. A. (2018). What Works and Does Not

Work in a Self-Management Intervention for People With Chronic Pain? Qualitative

Systematic Review and Meta-Synthesis. Physical Therapy, 98(5), 381–

397. https://doi.org/10.1093/ptj/pzy029

Dworkin, R. H., Turk, D. C., McDermott, M. P., Peirce-Sandner, S., Burke, L. B., Cowan, P., …

Sampaio, C. (2009). Interpreting the clinical importance of group differences in chronic

 15

pain clinical trials: IMMPACT recommendations. Pain, 146(3), 238–

244. doi:10.1016/j.pain.2009.08.019

Dysvik, E., Guttormsen Vinsnes, A., & Eikeland, O.-J. (2004). The effectiveness of a

multidisciplinary pain management programme managing chronic pain. International

Journal of Nursing Practice, 10(5), 224–234. doi:10.1111/j.1440-172x.2004.00485.x 

Dysvik, E., Kvaløy, J. T., Stokkeland, R., & Natvig, G. K. (2010). The effectiveness of a

multidisciplinary pain management programme managing chronic pain on pain

perceptions, health-related quality of life and stages of change--A non-randomized

controlled study. International journal of nursing studies, 47(7), 826–835.

https://doi.org/10.1016/j.ijnurstu.2009.12.001

Furnes, B., Natvig, G. K., & Dysvik, E. (2014). Therapeutic elements in a self-management

approach: experiences from group participation among people suffering from chronic

pain. Patient preference and adherence, 8, 1085–1092.

https://doi.org/10.2147/PPA.S68046

Biguet, G., Nilsson Wikmar, L., Bullington, J., Flink, B., & Löfgren, M. (2016). Meanings of

"acceptance" for patients with long-term pain when starting rehabilitation. Disability and

rehabilitation, 38(13), 1257–1267. https://doi.org/10.3109/09638288.2015.1076529

Glasziou, P., & Chalmers, I. (2018). Research waste is still a scandal—an essay by Paul Glasziou

and Iain Chalmers. BMJ, k4645. doi:10.1136/bmj.k4645 

Goldthorpe, J., Peters, S., Lovell, K., McGowan, L., & Aggarwal, V. (2016). 'I just wanted

someone to tell me it wasn't all in my mind and do something for me': Qualitative
 16

exploration of acceptability of a CBT based intervention to manage chronic orofacial

pain. British dental journal, 220(9), 459–463. https://doi.org/10.1038/sj.bdj.2016.332

Goodwin, K. A., & Goodwin, C. J. (2017). Research in psychology: Methods and designs (8th

ed.). Hoboken, NJ: John Wiley & Sons.

Helminen, E. E., Sinikallio, S. H., Valjakka, A. L., Väisänen-Rouvali, R. H., & Arokoski, J. P.

(2015). Effectiveness of a cognitive-behavioural group intervention for knee

osteoarthritis pain: a randomized controlled trial. Clinical rehabilitation, 29(9), 868–881.

https://doi.org/10.1177/0269215514558567

Joos, B., Uebelhart, D., Michel, B. A., & Sprott, H. (2004). Influence of an outpatient

multidisciplinary pain management program on the health-related quality of life and the

physical fitness of chronic pain patients. Journal of Negative Results in BioMedicine,

3(1). doi:10.1186/1477-5751-3-1

Monticone, M., Ambrosini, E., Rocca, B., Cazzaniga, D., Liquori, V., & Foti, C. (2016). Group-

based task-oriented exercises aimed at managing kinesiophobia improved disability in

chronic low back pain. European journal of pain (London, England), 20(4), 541–551.

https://doi.org/10.1002/ejp.756

Monticone, M., Ambrosini, E., Rocca, B., Cazzaniga, D., Liquori, V., Pedrocchi, A., & Vernon,

H. (2017). Group-based multimodal exercises integrated with cognitive-behavioural

therapy improve disability, pain and quality of life of subjects with chronic neck pain: a

randomized controlled trial with one-year follow-up. Clinical rehabilitation, 31(6), 742–

752. https://doi.org/10.1177/0269215516651979
 17

Nicholas, M. K., Asghari, A., Blyth, F. M., Wood, B. M., Murray, R., McCabe, R., Brnabic, A.,

Beeston, L., Corbett, M., Sherrington, C., & Overton, S. (2013). Self-management

intervention for chronic pain in older adults: a randomised controlled trial. Pain, 154(6),

824–835. https://doi.org/10.1016/j.pain.2013.02.009

Nordin, C., Gard, G., & Fjellman-Wiklund, A. (2013). Being in an exchange process:

experiences of patient participation in multimodal pain rehabilitation. Journal of

rehabilitation medicine, 45(6), 580–586. https://doi.org/10.2340/16501977-1136

Pincus, T., Anwar, S., McCracken, L. M., McGregor, A., Graham, L., Collinson, M., McBeth, J.,

Watson, P., Morley, S., Henderson, J., Farrin, A. J., & OBI Trial Management Team

(2015). Delivering an Optimised Behavioural Intervention (OBI) to people with low back

pain with high psychological risk; results and lessons learnt from a feasibility randomised

controlled trial of Contextual Cognitive Behavioural Therapy (CCBT) vs.

Physiotherapy. BMC musculoskeletal disorders, 16, 147. https://doi.org/10.1186/s12891-

015-0594-2

Rice AS, Smith BH, Blyth FM. Pain and the global burden of disease. Pain 2016;157(4):791-6.

Richards, L., & Morse, J. M. (2013). Readme first for a user's guide to qualitative methods (3rd

ed.). Los Angeles, CA: Sage Publications.

Toye, F., & Barker, K. (2012). 'I can't see any reason for stopping doing anything, but I might

have to do it differently'--restoring hope to patients with persistent non-specific low back

pain--a qualitative study. Disability and rehabilitation, 34(11), 894–903.

https://doi.org/10.3109/09638288.2011.626483
 18

Treede, R. D., Rief, W., Barke, A., Aziz, Q., Bennett, M. I., Benoliel, R., Cohen, M., Evers, S.,

Finnerup, N. B., First, M. B., Giamberardino, M. A., Kaasa, S., Kosek, E.,

Lavandʼhomme, P., Nicholas, M., Perrot, S., Scholz, J., Schug, S., Smith, B. H.,

Svensson, P., … Wang, S. J. (2015). A classification of chronic pain for ICD-

11. Pain, 156(6), 1003–1007. https://doi.org/10.1097/j.pain.0000000000000160

Williams, A., Fisher, E., Hearn, L., & Eccleston, C. (2020). Psychological therapies for the

management of chronic pain (excluding headache) in adults. The Cochrane database of

systematic reviews, 8(8), CD007407. https://doi.org/10.1002/14651858.CD007407.pub4