Key messages

• Skill-mix interventions involving pharmacists are effective at improving blood glucose levels and blood pressure control and have comparable results on health-related quality of life.

• Pharmacist-delivered care can improve medication adherence for patients with a variety of chronic conditions.

• There is some, albeit limited, evidence showing that skill-mix interventions involving pharmacists may be cost-effective and reduce the risk of hospital admissions.

Chronic conditions

Evidence on health and profession-specific outcomes

With respect to diabetes, included reviews provide strong evidence that pharmacist-involved interventions (for example, more advanced roles in patient education, medication and disease management) compared with usual care can significantly improve glycated haemoglobin (HbA1c) levels, irrespective of whether pharmacists work as part of a multiprofessional team or autonomously (Table 6.1) (Reference Aguiar, Brito G de and Lima TdeAguiar et al., 2016; Reference Armor, Britton and DennisArmor et al., 2010; Reference Deters, Laven and CastejonDeters et al., 2018; Reference Fazel, Bagalagel and LeeFazel et al., 2017; Greer et al., 2016; Reference Nkansah, Mostovetsky and YuNkansah et al., 2010; Reference van Eikenhorst, Taxis and van Dijkvan Eikenhorst et al., 2017).

The evidence also suggests that pharmacist-delivered care improves cardiovascular disease outcomes, with studies reporting improvements in systolic and diastolic blood pressure measures versus usual care (Reference Blalock, Roberts and LauffenburgerBlalock et al., 2013; Reference Cheema, Sutcliffe and SingerCheema, Sutcliffe & Singer, 2014; Reference Fazel, Bagalagel and LeeFazel et al., 2017; Greer et al., 2016; Reference Morgado, Morgado and MendesMorgado et al., 2011; Reference Nkansah, Mostovetsky and YuNkansah et al., 2010; Reference van Eikenhorst, Taxis and van Dijkvan Eikenhorst et al., 2017). Pharmacist-delivered care had comparable effects on health-related quality of life and mortality risk for patients on warfarin compared with usual physician-delivered care (Reference Entezari-Maleki, Dousti and HamishehkarEntezari-Maleki et al., 2016). Some evidence suggested pharmacist interventions improved quality of life for patients with asthma, heart failure and those at high risk of medications related problems (Reference Nkansah, Mostovetsky and YuNkansah et al., 2010). Evidence also indicated a lower risk of major bleeding and thromboembolic events for participants engaged in pharmacist-delivered interventions, albeit with significance only found in non-RCT studies (Reference Entezari-Maleki, Dousti and HamishehkarEntezari-Maleki et al., 2016; Reference Manzoor, Cheng and LeeManzoor et al., 2017).

Pharmacist-delivered care for patients living with cancer led to improved control of nausea and vomiting, patient satisfaction, medication adherence and increased knowledge of blood pressure management. It was also shown to improve quality of life and patient satisfaction for adults on antineoplastic medication (Reference Colombo, Aguiar and LimaColombo et al., 2017). However, no significant effect of pharmacist-delivered care was reported in one study on patients with chronic kidney disease for quality of life, renal outcomes and blood pressure (Reference Nicoll, Robertson and MarksNicoll et al., 2018), nor in one study on COPD care for lung function (Reference Zhong, Ni and CuiZhong et al., 2014). In addition, one review concluded that pharmacist-delivered care did not improve depression symptoms (Reference Readdean, Heuer and Scott ParrottReaddean, Heuer & Scott Parrott, 2018).

Compared with usual care, pharmacist-delivered interventions were able to significantly improve medication adherence for patients with hypertension (Reference Blalock, Roberts and LauffenburgerBlalock et al., 2013; Reference Cheema, Sutcliffe and SingerCheema, Sutcliffe & Singer, 2014; Reference Morgado, Morgado and MendesMorgado et al., 2011), diabetes (Reference Blalock, Roberts and LauffenburgerBlalock et al., 2013; Reference Fazel, Bagalagel and LeeFazel et al., 2017; Reference van Eikenhorst, Taxis and van Dijkvan Eikenhorst et al., 2017; Reference Wang, Yeo and KoWang, Yeo & Ko, 2016), depression (Reference Readdean, Heuer and Scott ParrottReaddean, Heuer & Scott Parrott, 2018) and COPD (Reference Zhong, Ni and CuiZhong et al., 2014). One review on pharmacist-delivered patient care services in the community for people with various chronic conditions or multimorbidity concluded that it improved patient satisfaction but did not reduce safety outcomes or quality of life (Reference Blalock, Roberts and LauffenburgerBlalock et al., 2013).

With respect to profession-specific outcomes, limited evidence also suggests that pharmacist interventions can improve adherence to prescribing guidelines for physicians, resulting in a reduction in therapeutic duplication and the number of medications prescribed (Reference Nkansah, Mostovetsky and YuNkansah et al., 2010). Two studies also reported that pharmacist-involved care increased adherence to American Diabetes Association standards of care for most care types (Reference Armor, Britton and DennisArmor et al., 2010) and led to a higher rate of prescribing of vitamin D and bicarbonate for patients with chronic kidney disease (Reference Nicoll, Robertson and MarksNicoll et al., 2018).

Evidence on health-system-related outcomes

Five studies reported economic outcomes on pharmacist-involved interventions. According to one review (Reference Wang, Yeo and KoWang, Yeo & Ko, 2016), pharmacist-managed care resulted in higher quality-adjusted life-years with lower costs than usual care (so the intervention was dominant). Pharmacist-managed warfarin services were shown to significantly lower emergency department visits and hospitalization rates and to be cost saving over usual physician-delivered care (Reference Entezari-Maleki, Dousti and HamishehkarEntezari-Maleki et al., 2016; Reference Manzoor, Cheng and LeeManzoor et al., 2017). Pharmacist-delivered care was also shown to lower anti-emetic drug costs for cancer care (Reference Colombo, Aguiar and LimaColombo et al., 2017), while lower costs were reported for collaborative care models involving pharmacists and pharmacist-provided consultations for high-risk diabetes patients (Reference Armor, Britton and DennisArmor et al., 2010) and for COPD patients (Reference Zhong, Ni and CuiZhong et al., 2014).

Pharmacist-directed care and pharmacist-collaborative care were also found to significantly reduce risk of hospital admissions for patients with COPD, but had no effect on emergency department visits (Reference Zhong, Ni and CuiZhong et al., 2014). One review also found similar rates of emergency department visits, office visits and hospitalizations compared with usual care for people with chronic conditions (Greer et al., 2016). There was, however, some evidence of pharmacist-delivered care leading to a higher percentage of asthma patients with breathing-related emergency department visits or hospitalizations (Reference Blalock, Roberts and LauffenburgerBlalock et al., 2013).

Multimorbidity

The one review evaluating care for multimorbidity involving pharmacists, focused on pharmacist-directed care and pharmacist-collaborative care with primary health practitioners. This review found evidence that multiprofessional teamwork significantly improved systolic and diastolic blood pressure, low-density lipoprotein cholesterol (but not high-density lipoprotein cholesterol), HbA1c levels and 10-year Framingham risk score versus usual care (Reference Tan, Stewart and ElliottTan et al., 2014).

Key messages

• Nurse-delivered care leads to equivalent or better health outcomes compared with the usual care, in particular titration of medication by nurses and nurse-led clinics.

• There is insufficient evidence on the impact of skill-mix interventions involving nurses in expanded roles on health system outcomes and cost effectiveness, but some studies show a positive impact on hospital (re)admissions, physician visits and costs.

• Evidence on the impact of skill-mix innovations involving nurses on health outcomes for people with multimorbidity is inconclusive.

Chronic conditions

Evidence on health and profession-specific outcomes

The included reviews focusing on diabetes found a comparable effect on HbA1c levels for independent nurse prescribing compared with physician-led prescribing (Health Quality Ontario, 2013;

Table 6.2 is available online at www.cambridge.org/skill-mix

Reference Martínez-González, Rosemann, Tandjung and DjalaliMartínez-González et al., 2015; Reference Tabesh, Magliano and KoyeTabesh et al., 2018). Moreover, titration of medication by nurses significantly reduced HbA1c levels and improved health behaviours and medication adherence (Reference Shaw, McDuffie and HendrixShaw et al., 2013). Self-management education by nurses was nevertheless shown to have no significant impact on HbA1c levels (Reference Parker, Maresco-Pennisi and CliftonParker et al., 2016), total mortality, quality of life or cholesterol levels (Reference Massimi, De Vito and BrufolaMassimi et al., 2017).

With respect to cardiovascular disease, nurse-titration of medication was reported to significantly reduce both systolic and diastolic blood pressure (Reference Parker, Maresco-Pennisi and CliftonParker et al., 2016; Reference Shaw, McDuffie and HendrixShaw et al., 2013). In addition, nurse-led cardiac clinics were found to provide equivalent or better care than clinics run by other health professionals or usual care of cardiovascular disease. Compared with usual care, studies reported significant reductions in total mortality and equivalent outcomes for self-reported mental or physical health compared with non-nurse led clinics (Reference Schadewaldt and SchultzSchadewaldt and Schultz, 2010). Nurse-delivered motivational interviewing was shown to increase medication adherence for patients with cardiovascular disease (Reference Al-Ganmi, Perry and GholizadehAl-Ganmi et al., 2016).

Moderate evidence also suggests that nurse-management of antidepressant medication adherence programmes, with nurses working independently or as part of a multiprofessional team, helps to lower depression symptom severity and significantly improved patient satisfaction, but may not improve medication adherence (Reference Heise and van ServellenHeise & van Servellen, 2014). Limited evidence indicates that self-management education delivered by nurses, peers or health assistants can lead to small improvements in physical health and increased use of primary care services for people with a severe mental health condition (Reference Kelly, Fenwick and BarrKelly et al., 2014).

One review assessed psychosocial interventions for cancer patients and their families delivered predominantly by nurses (or psychotherapists and social worker) (Reference Hopkinson, Brown and OkamotoHopkinson et al., 2012). The review found some evidence of improved depression and anxiety symptoms and improved quality of life for cancer patients and improved patient–carer relationships. Nevertheless, there was also some evidence of worse psychosocial status in carers taught to give direct care and adverse consequences for the health status of cancer patients, including increased depression in those receiving cognitive behavioural therapy (Reference Hopkinson, Brown and OkamotoHopkinson et al., 2012).

Nurse-delivered care was found to lead to equivalent or better outcomes than usual care for patients with chronic kidney disease. Reviews reported significant improvement of symptoms and sleep, equivalent outcomes for patient satisfaction, mortality and renal outcomes (Reference Chen, Chen and LiuChen et al., 2016; Reference Nicoll, Robertson and MarksNicoll et al., 2018).

Four reviews assessed nurse-delivered interventions for patients with respiratory diseases. Nurse-delivered self-management for patients with COPD was shown to significantly reduce anxiety symptoms (Reference Baishnab and KarnerBaishnab et al., 2012) and nurse-led asthma clinics had a significant effect on nocturnal awakenings (Reference Baker and FatoyeBaker et al., 2017). For asthma patients, there was comparable effect on number of exacerbations (Reference Baker and FatoyeBaker et al., 2017; Reference Kuethe, Vaessen-Verberne and ElbersKuethe et al., 2013), quality of life, disease severity and symptoms as well as use of preventer medication (Reference Baishnab and KarnerBaishnab et al., 2012; Reference Kuethe, Vaessen-Verberne and ElbersKuethe et al., 2013) and levels of correct inhalation technique and well-controlled asthma (Reference Martínez-González, Rosemann, Tandjung and DjalaliMartínez-González et al., 2015). Nurse-delivered home care and education for patients with obstructive sleep apnoea was shown to have no significant effect on continuous positive airway pressure and physical functioning compared with physician-delivered care (Reference Gong, Chen and WuGong et al., 2018).

For patients with Parkinson’s disease, nurse-delivered autonomous care was shown to significantly improve best hand score compared with usual care (Reference Martínez-González, Rosemann, Tandjung and DjalaliMartínez-González et al., 2015), while nurse practitioner- and specialist nurse-delivered care within a team significantly increased the proportion of individuals with a chronic disease appropriately receiving influenza or pneumovax vaccinations as well as patient education related to smoking, exercise, diet, and medication side effects (Health Quality Ontario, 2013).

Four studies reported on professional-specific outcomes. For patients with various chronic conditions, nurse practitioner- or specialist nurse-delivered clinical care within a multiprofessional team was shown to lead to no significant difference in job satisfaction, or inappropriate demands for teamwork (Health Quality Ontario, 2013), while no difference in adherence to practical guidelines was found for nurse-delivered autonomous care (Reference Martínez-González, Rosemann, Tandjung and DjalaliMartínez-González et al., 2015). However, nurse-led care for patients with chronic kidney disease was shown to lead to improvements in staff encouragement and an increase in prescribing of relevant drugs (Reference Chen, Chen and LiuChen et al., 2016; Reference Nicoll, Robertson and MarksNicoll et al., 2018). Limited evidence also suggests that nurse practitioner- or specialist nurse-delivered clinical care increased the number of referrals for echocardiographs among patients with presumed coronary heart failure and the assessment of blood pressure, smoking status and body mass index/weight among coronary artery disease patients (Health Quality Ontario, 2013). Nurse practitioners and specialist nurses were also more likely to intensify glucose-lowering therapy, intensify blood pressure medications or refer patients with diabetes to an internist for the start of insulin therapy (Health Quality Ontario, 2013).

Evidence on health-system-related outcomes

Ten reviews reported on health-system-related outcomes of which most outcome measures covered the use of health resources. There is some evidence that nurse-led collaborative care can reduce hospitalizations for patients with coronary artery disease (Health Quality Ontario, 2013). Nurses providing interventions to improve the transition from inpatient to outpatient care were found to reduce readmission among adults with mental illness (Reference Vigod, Kurdyak and DennisVigod et al., 2013). Nurse-delivered care in general was reported to improve secondary prevention of heart disease by appropriately increasing aspirin intake and adoption of low-fat diets and physically activity (Reference Martínez-González, Rosemann, Tandjung and DjalaliMartínez-González et al., 2015). Furthermore, nurse-delivered care for patients with chronic kidney disease significantly reduced the number of physician visits, associated with an increased number of nurse visits (Reference Nicoll, Robertson and MarksNicoll et al., 2018). However, no significant difference was reported on accident and emergency department attendance and hospital admissions for patients attending nurse-led asthma and cardiac clinics (Reference Baishnab and KarnerBaishnab et al., 2012; Reference Kuethe, Vaessen-Verberne and ElbersKuethe et al., 2013; Reference Schadewaldt and SchultzSchadewaldt and Schultz, 2010).

In terms of economic outcomes, nurse or pharmacist-delivered care in the home was shown to improve self-management of cardiovascular disease, reduce all-cause mortality and hospitalizations, resulting in average cost savings over usual care of $10 665 per patient (Reference Fergenbaum, Bermingham and KrahnFergenbaum et al., 2015). Nurse-delivered asthma management also led to significantly lower costs for outpatient visits (Reference Kuethe, Vaessen-Verberne and ElbersKuethe et al., 2013), and nurse titration of medicines following a protocol led to lower inpatient costs but not significant differences in outpatient costs for patients at risk of cardiovascular disease (Reference Shaw, McDuffie and HendrixShaw et al., 2013). Limited evidence suggests that titration of medication by nurses lowers inpatient costs as well as salary costs compared with usual care (Reference Shaw, McDuffie and HendrixShaw et al., 2013).

Multimorbidity

Of the nine reviews on multimorbidity, four reviews analysed the expansion of roles for nurses and found improved or equivalent patient outcomes compared with usual care or physician-delivered care models (Table 6.5). Nurse-prescribing and community monitoring moderately reduced systolic but not diastolic blood pressure compared with physician-delivered care, while nurse-led clinics provided equivalent care to other clinics (Reference Clark, Smith and TaylorClark et al., 2011). All care models had a comparable effect on medication adherence compared with physician-delivered care (Reference Clark, Smith and TaylorClark et al., 2011). Nurse-delivered education to improve self-monitoring and disease management was shown to significantly reduce systolic and diastolic blood pressure and HbA1c levels, with a weak trend towards lower mortality (Reference Massimi, De Vito and BrufolaMassimi et al., 2017). Nurse-led collaborative care of patients significantly reduced depression severity for patients with depression and a physical health condition (Reference Ekers, Murphy and ArcherEkers et al., 2013), with limited evidence for improvements on patient, caregiver and health care staff satisfaction. No impact was observed on emergency admissions and bed days (Reference Lupari, Coates and AdamsonLupari et al., 2011).

Key messages

• Peer educators and community health workers – if involving regular contact with patients – were shown to improve several health outcomes and peer-related outcomes and/or to have comparable effects.

• The evidence is mixed as to whether transition coordinators (by various health professions or multiprofessional teams) can improve certain health outcomes.

• Disease management delivered by different professions or multiprofessional teams was shown to improve HbA1c levels, self-reported health status and patient satisfaction.

• Skill-mix interventions involving patient navigators, transition coordinators or peers reported a positive impact on resource use for cancer and diabetes patients and patients with mental illness. However, evidence on impact of skill-mix interventions on resource use overall is mixed and remains insufficient.

Evidence on health and profession-specific outcomes

Diabetes disease management delivered by different professions or multiprofessional teams significantly reduced HbA1c levels but had no impact on mortality risk (Reference Pimouguet, Le Goff and ThiébautPimouguet et al., 2011). The transition coordinator model did not lead to overall improvements in HbA1c levels (Reference Chu, Maslow and IsenburgChu et al., 2015).

Peer support (generally lay persons with experience of the same disease(s) or community health workers) was found to significantly reduce HbA1c levels compared with usual care, provided that there was a moderate to high frequency of contact with patients (Reference Qi, Liu and QiQi et al., 2015). Peer support was also shown to significantly improve blood pressure control, cholesterol, body mass index, physical activity levels, self-efficacy, depression and perceived social support (Reference Dale, Williams and BowyerDale, Williams & Bowyer, 2012). However, peers involved in case management for adults with severe mental illness had no significant effect on symptoms, patient satisfaction and quality of life (Reference Wright-Berryman, McGuire and SalyersWright-Berryman, McGuire & Salyers, 2011).

With respect to depression, training for primary care practitioners failed to improve clinical symptoms unless implemented alongside provision of specific care guidelines (Reference Sikorski, Luppa and KönigSikorski et al., 2012). Training for parents to better manage children with autism spectrum disorder had no impact on comprehension, child initiation, expression and joint language but showed significant improvement of severity of autism characteristics, shared attention in parent–child interaction as well as parent satisfaction (Reference Oono, Honey and ConachieOono, Honey & McConachie, 2013).

Physiotherapists providing management of musculoskeletal disorders were shown to improve patient satisfaction compared with physician care, and had comparable effects on general health outcomes (Reference Marks, Comans and BissetMarks et al., 2017). Self-management education provided by various health professionals for asthma patients was shown to have comparable effects on asthma symptoms, quality of life scores and number of school days missed (Reference Welsh, Hasan and LiWelsh, Hasan & Li, 2011). One review looked at case management carried out by various professionals either independently or within multiprofessional teams for patients with various chronic conditions (mostly frail elderly); the review found significant improvements of self-reported health status in the short-term and patient satisfaction in the long-term and comparable effects on mortality (Reference Stokes, Panagioti and AlamStokes et al., 2015).

Evidence on health-system-related outcomes

For cancer care, GP-led follow-up care for child cancer survivors delivered autonomously or collaboratively with paediatric cancer specialists was reported to reduce patient travel and attrition for follow-up appointments, with some evidence of improved cost-effectiveness compared with follow up in long-term follow-up clinics (Reference Singer, Gianinazzi and HohnSinger et al., 2013). Nevertheless, there was also evidence of reduced access to complex diagnostic testing compared with follow up in specialist care settings (Reference Singer, Gianinazzi and HohnSinger et al., 2013). Patient navigation for breast cancer care involving lay workers and peers was reported to improve adherence to screening and diagnostic follow up after abnormal radiographic screening and to reduce waiting times for appointments with genetic counsellors and for biopsy/diagnostic intervals (Reference Robinson-White, Conroy and SlavishRobinson-White et al., 2010).

With respect to diabetes care, the transition coordinator and community health worker model produced higher rates of successful transfer and attendance in adult clinics compared with usual care (Reference Chu, Maslow and IsenburgChu et al., 2015) and a decrease of medication use and patient costs (Reference Jack, Arabadjis and SunJack et al., 2016).

For musculoskeletal disorders, physiotherapists made similar diagnostic and management decisions to specialists. However, results on resource utilization are inconsistent and evidence on costs is lacking (Reference Marks, Comans and BissetMarks et al., 2017). Consumer–provider models in which mental health services are provided by peers improved patient engagement in treatment and social relationships but studies showed limited support for reduced hospitalizations (Reference Wright-Berryman, McGuire and SalyersWright-Berryman, McGuire & Salyers, 2011). For self-management education provided by various health professionals for asthma patients mixed evidence was found on hospital admissions and no effect on emergency department visits (Reference Welsh, Hasan and LiWelsh, Hasan & Li, 2011).

Skill-mix interventions delivered by multiprofessional teams

Key messages

• Collaborative care and interdisciplinary care were shown to have a moderate impact on physical health (HbA1c levels, blood pressure, progression of chronic kidney disease, physical functioning), medication adherence and patient satisfaction, and may improve mental health symptoms and quality of life.

• Other multiprofessional care models (for example, primary care networks, multiprofessional clinics) led to few improvements in physical or mental health (of which included reduction in all-cause mortality) but reduced utilization of inpatient care and other health care services.

• Specialty-based shared multiprofessional care for HIV patients saw positive effects on health outcomes and treatment adherence, though it is limited around whether it is cost-effective.

Collaborative care models for people with comorbidities improved adherence to medication and some health outcomes (depression, Quality-Adjusted Life-Years), but showed limited evidence on cost savings.

Chronic conditions

Collaborative care models and consultation liaison

Collaborative care between primary care physicians and specialists had a modest impact on reducing HbA1c levels, blood pressure outcomes and medication adherence compared with usual care in two studies (Foy et al., 2010a; Reference Mitchell, Burridge and ZhangMitchell et al., 2015), although one study found no significant impact (Reference Smith, Cousins and ClyneSmith et al., 2017). Collaborative care did, however, significantly improve physical functioning, patient satisfaction and clinical attendance rates at a moderately increased cost compared with usual care for treatment of heart failure (Reference Mitchell, Burridge and ZhangMitchell et al., 2015). Multiprofessional care models involving various health professionals were shown to have a significant effect in delaying progression of chronic kidney disease (Reference Strand and ParkerStrand & Parker, 2012) and on the number of patients with hypertension, with no difference reported in renal outcomes and composite cardiovascular end-points (Reference Nicoll, Robertson and MarksNicoll et al., 2018).

Strong evidence suggests that collaborative care involving a mental health specialist and primary care physicians/providers significantly improved depressive symptoms (Reference Archer, Bower and GilbodyArcher et al., 2012; Reference Coventry, Hudson and KontopantelisCoventry et al., 2014; Reference Foy, Hempel and RubensteinFoy et al., 2010; Reference Sighinolfi, Nespeca and MenchettiSighinolfi et al., 2014; Reference Smith, Cousins and ClyneSmith et al., 2017; Reference Thota, Sipe and ByardThota et al., 2012), patient satisfaction, quality of life and medication adherence versus usual care (Reference Archer, Bower and GilbodyArcher et al., 2012; Reference Coventry, Hudson and KontopantelisCoventry et al., 2014; Reference Thota, Sipe and ByardThota et al., 2012). Similarly, consultation liaison between primary care providers (mostly GPs) and mental health specialists was shown to improve adherence to treatment guidelines, consumer satisfaction and delivery of adequate treatment from primary care practitioners (Reference Gillies, Buykx and ParkerGillies et al., 2015). Consultation liaison was also found to improve mental health in the short-term and to be similar to usual care at 12 months follow up, with no differences reported in the number of health care visits. There was, however, some limited evidence of higher costs over usual care (Reference Gillies, Buykx and ParkerGillies et al., 2015).

Multiprofessional care coordination and intensive case management

There is equivocal evidence on the impact of multiprofessional care coordination. Reviews on cancer care suggest that coordination of care failed to significantly improve functional status, physical or mental health or quality of life (Reference Aubin, Giguère and MartinAubin et al., 2012; Reference Gorin, Haggstrom and HanGorin et al., 2017) or continuity of care (Reference Aubin, Giguère and MartinAubin et al., 2012). It may nevertheless enhance patient satisfaction, quality of end-of-life care and the odds of appropriate health care utilization in primary, acute and hospice care settings (Reference Gorin, Haggstrom and HanGorin et al., 2017).

Findings from one review suggest that intensive case management does not lead to significant improvements for mental health conditions, mortality risk, social functioning or quality of life, although it did lead to a slight reduction in length of hospital stay for adults with severe mental illness (Reference Dieterich, Irving and ParkDieterich et al., 2010). Another review reported that intensive case management did improve quality of life and social contact and reduce distress behaviour and caregiver burden in older adults with dementia (Abendstern et al., 2012).

Other multiprofessional care models

Primary care networks and family medicine groups for diabetes care did not improve self-reported physical and mental health, but were shown to reduce emergency department and hospital visits (Reference Carter, Riverin and LevesqueCarter et al., 2016). Multiprofessional cardiac clinics led to fewer hospitalizations and a significant reduction in all-cause mortality (Reference Gandhi, Mosleh and SharmaGandhi et al., 2017).

Studies on mental health reported that consumer–provider models where mental health service users worked in partnership with mental health professionals did not improve mental health or client satisfaction outcomes, but led to small reductions in the use of crisis and emergency services (Reference Pitt, Lowe and HillPitt et al., 2013). Collaborative care interventions in which a designated psychiatrist and a nurse care coordinator worked together to address physical and mental health needs of people with severe mental illness demonstrated significantly lower hospitalization rates for patients, but only for the second year of the 3-year follow-up period. In comparison to the standard care group, the intervention significantly improved the Mental Health Component of quality of life but not the Physical Health Component of this measure (Reference Reilly, Planner and GaskReilly et al., 2013).

Shared care models and other specialty-based multiprofessional care for HIV-positive individuals showed positive effects on treatment adherence, infection and transmission rates and mortality (Reference Wong, Luk and KiddWong, Luk & Kidd, 2012; Reference Kimmel, Martin and GaladimaKimmel et al., 2017). Shared care models with primary care clinicians seem also to be more cost-effective for HIV screening compared with specialist care (Reference Wong, Luk and KiddWong, Luk & Kidd, 2012). However, evidence is limited to determine whether HIV shared care is cost-effective (Reference Mapp, Hutchinson and EstcourtMapp, Hutchinson & Estcourt, 2015).

Multimorbidity

Collaborative care models for people with comorbid depression and diabetes or cardiovascular disease were shown to improve depression outcomes (Reference Atlantis, Fahey and FosterAtlantis, Fahey & Foster, 2014; Reference Huang, Wei and WuHuang et al., 2013; Reference Jeeva, Dickens and CoventryJeeva et al., 2013), reduce HbA1c levels (Reference Atlantis, Fahey and FosterAtlantis, Fahey & Foster, 2014; Reference Huang, Wei and WuHuang et al., 2013) and improve quality-adjusted life-years (Reference Huang, Wei and WuHuang et al., 2013; Reference Jeeva, Dickens and CoventryJeeva et al., 2013) and also to lead to higher rates of adherence to antidepressant medication and to oral hypoglycaemic agent (Reference Huang, Wei and WuHuang et al., 2013). No statistically significant cost savings between intervention and usual care groups were identified (Reference Jeeva, Dickens and CoventryJeeva et al., 2013).

Patient-focused education and collaborative care by nurses, pharmacists and social workers for people with multimorbidity improved prescribing and drug adherence, although there was little evidence of improved health outcomes, reduced health care utilization or cost savings (Reference Smith, Soubhi and FortinSmith et al., 2012).

Collaborative goal-setting between older patients with complex multimorbidity and a primary care physician within an integrated care strategy significantly improved goal-setting, the number of advance directives and the inclusion of goals in care plans for multifactorial, but not for single-factorial, interventions (Reference Vermunt, Harmsen and WestertVermunt et al., 2017).

Strengths and limitations of the evidence

Evidence from the overview of reviews reveals that several skill-mix models (expanded roles for nurses and pharmacists and, based on weaker evidence, other professional roles and multiprofessional teams) have overall positive impacts on health outcomes for patients with chronic conditions, but uncertain effects on economic benefits and quality of care.

Allocating tasks such as patient education and health screenings to pharmacists was shown to effectively improve blood glucose levels, blood pressure control and medication adherence, with outcomes for chronic kidney disease, COPD and depression similar to those for usual care. There was also some evidence suggesting that pharmacist-involved care could improve professional-specific outcomes, most notably in terms of enhancing physician adherence to prescribing or other care guidelines. Included reviews also provided limited evidence that pharmacist-delivered care for diabetes, cardiovascular disease, hypertension and COPD could be cost saving compared with physician- and nurse-delivered care, with mixed evidence showing pharmacist interventions either improved the rate, or led to a similar rate, of health care utilization compared with usual care.

Nurse-delivered autonomous care including prescribing, titration of medication and nurse-led clinics also provided equivalent or better care compared with that provided by physicians or other health professionals, although nurse-led collaborative care in general did not improve health outcomes. There is insufficient evidence to make any assessments on the cost-effectiveness of skill-mix interventions involving nurses or their impact on health care utilization.

Evidence does not strongly indicate that multiprofessional care models lead to better physical health outcomes. However, it suggests that interdisciplinary care coordination, especially between primary and secondary care, can significantly improve mental health outcomes for patients with severe mental health issues or multimorbidity and may increase patient satisfaction.

Overall, the evidence base was considerable, as shown by a large number of reviews. Moreover, 43 reviews conducted a meta-analysis and sufficiently accounted for biases and heterogeneity in original studies when interpreting results. Nonetheless, there are a number of limitations in the evidence base. First, there was limited evidence on skill-mix interventions for the management of multimorbidity care. Second, the description of comparison groups was often sparse, and the education/training of professionals was generally not described. Third, although there was good evidence on health outcomes, few reported on other relevant outcomes such as health care utilization, economic benefits or professional-related outcomes. Third, systematic reviews were dominated by studies from the USA, the United Kingdom, Australia and Canada. As the impact of skill-mix interventions is likely to differ depending on specific geographical and cultural contexts and health system models, the generalizability and transferability of results may be limited. Similarly, and lastly, the contexts in which skill-mix initiatives were implemented and in which studies were undertaken differ across the included reviews and generalized conclusions must be made cautiously.

Overall, more research is needed to assess a wider spectrum of outcomes in more countries to determine whether skill-mix interventions can cost-effectively improve patient outcomes and quality of care for patients with single chronic conditions or multimorbidity.

The role expansion of nurses and other health professionals for the management of patients with chronic conditions

New forms of division of work and re-allocation of responsibilities have emerged across Europe, involving nurses in the management of patients with chronic conditions. The increases in patient volumes and complexity have led to changes in the division of work. In these models, physicians often take care of the more complex patients, whereas stable chronic patients are treated by nurses or other professions. Examples include task shifting and sharing between physicians and other professions, such as nurses, pharmacists, physician assistants or medical assistants, which is facilitated by expanding their scopes-of-practice. This follows a global trend to expand the role of nurses for specific, chronic conditions in primary care. For example, reforms have been introduced in Australia, Canada, Finland, Ireland, the Netherlands, New Zealand, the United Kingdom and the USA to widen the scope of practice of nurse practitioners and other advanced practice nursesFootnote ² in the care of patients with chronic conditions (Maier, Aiken & Busse, 2017). Similarly, a new profession, physician assistants, has expanded globally, but they are numerically smaller and work less often in primary care than nurses or pharmacists (Reference Hooker and EverettHooker & Everett, 2012).

Nurses’ roles are often broadened when medical shortages are anticipated. In France, skill-mix and professional cooperation initiatives were launched with support from key stakeholders in the face of such shortages between 2004 and 2008 (Reference Mousquès, Bourgueil and Le FurMousquès et al., 2010). One was the ASALEE project (French acronym for Health Action by Teams of Self-employed Health Professionals) that was initiated in 2004 as a non-profit organization of several GP practices in a rural area of France to improve care, especially in the most remote or deprived areas, for patients with chronic conditions and multimorbidity. It employed nurses directly to collaborate closely with two to three independently practising GPs, after specific training. The nurses were delegated tasks in prevention and chronic disease management that were previously performed by GPs (such as therapeutic education consultations for diabetes and high blood pressure, cognition and cardiovascular risk factor screenings for individuals aged over 75 years and computerized management of patient data) (Reference Supper, Bourgueil and EcochardSupper et al., 2017; Bourgueil et al., forthcoming). The project has since expanded to all regions of France, involving about 2% of all practising GPs, 300 full-time equivalent nurses and 300 000 patients. Initially funded by innovation funds, ASALEE is now financed by national insurance, conditional upon economic evaluation since 2017 (Bourgueil et al., forthcoming). Evaluations of the management of type 2 diabetes showed improved glycaemic control and follow up of patients treated within ASALEE, without additional cost (Bourgeuil et al., 2008).

An important, related reform trend is the introduction of laws on nonmedical prescribing. Between 2010 and 2016 alone, seven countries (Cyprus, Estonia, Finland, France, the Netherlands, Poland and Spain) adopted laws that grant specialized nurses (nurse specialists, clinical nurse specialists or professional nurses with additional education) the authority to prescribe medication (Maier, Aiken & Busse, 2017). Ireland, the United Kingdom, Denmark and Sweden had authorized limited nurse prescribing before 2010. Since 2012, nurse specialists in the Netherlands with a 2-year Master’s degree in Advanced Nursing Practice can prescribe licensed medicines for medical conditions within their specialty area. Since 2014, registered nurses with a Bachelor’s degree and a completed pharmacotherapy module in diabetes, lung and oncology care may prescribe a limited number of prescription-only medications after diagnosis by a physician (Reference KroezenKroezen, 2014). In Finland, where nurses are authorized to prescribe a limited set of medicines since 2010, the shifting of prescribing rights to nurses is less extensive than in the Netherlands, but with considerable change in the division of responsibility, in particular for patients with chronic conditions, as they can receive prescriptions and follow-up care from nurse prescribers in health centres (see Box 6.7, below).

Pharmacists are increasingly involved in the management of patients with chronic conditions

The role of pharmacists working in community pharmacies has substantially expanded over the last 10 years. As community-based accessible health care providers, pharmacists are well positioned to provide low threshold clinical and counselling services and drug therapy expertise to local populations. Their role is not only promising regarding medication management but also for disease management, including measurement and testing of clinical parameters, support in self-care/management-related interventions (Pharmaceutical Group of The European Union, 2017), structured patient education and adherence improvement counselling.

Services and activities provided by pharmacists are mostly provided at national level and are defined either in law or specific government policy. Pharmacies led the management of common chronic diseases such as asthma/COPD, diabetes and hypertension in more than one third of 30 European countries in 2016. Also, the share of pharmacies offering measurement and testing services for chronic conditions (blood pressure, weighing, hypertension, cholesterol) has significantly increased over the last 5 years across Europe. In 2012, about half of 30 European countries’ pharmacies offered blood pressure measurement and weighing; in 2016 these services were offered in about 90% of the countries (Pharmaceutical Group of The European Union, 2013, 2017). Case study 2 describes how the scope of pharmacists’ tasks in Portugal has grown substantially in the last decade.

Disease management as a single-disease integrated care model

Around 2000, many European countries began implementing DMPs as a way to effectively and efficiently improve the management of chronic disease care. DMPs are designed to create a continuum of care for single chronic diseases (mostly diabetes, cardiovascular disease and COPD), provided by a multiprofessional team of health care professionals, and to encourage patient self-management (Reference Rijken and BekkemaRijken and Bekkema, 2011). Although in several countries, GPs direct the care processes within a DMP, care managers (usually nurses or allied health care professionals) are increasingly appointed to assume the coordinating tasks. Called nurse-delivered or nurse-led DMPs, these have added to the types and implementation of collaborative care models and care coordination (see also next section).

Moreover, several skill-mix changes implemented as part of integrated chronic care models demonstrate the interrelatedness of skill-mix interventions and organizational re-design as was shown by the EU-funded INTEGRATE project. The skill-mix innovations that are frequently implemented in combination include nurse involvement in care delivery, multiprofessional teams and protocols/pathways, continuous training of health professionals, involvement of a case manager/care coordinator role, regular team meetings to discuss a patient’s treatment, and the creation of a new position, role or function specifically to deliver integrated chronic care (Reference Busetto, Luijkx and CalciolariBusetto et al., 2017).

Collaboration and multiprofessional team-based care

Many European countries have changed their chronic disease management strategies over the last decade as DMPs were increasingly criticized as inappropriate to meet the needs of the growing numbers of patients with multiple chronic conditions (Reference Nolte, Knai and SaltmanNolte, Knai & Saltman, 2014; Reference Rijken, Bekkema and BoeckxstaensRijken et al., 2014; Reference Starfield, Nolte and McKeeStarfield, 2008). Care for multimorbidity comprises specialized care, intersectoral coordination and collaboration between primary care and specialized care (Reference Hujala, Taskinen and RissanenHujala, Taskinen & Rissanen, 2017; Reference Rijken, Hujala and GinnekenRijken et al., 2018; Reference Struckmann, Quentin and BusseStruckmann et al., 2017). To better respond to the needs of patients with multiple chronic conditions, countries are moving towards more comprehensive models of integrated care by looking beyond disease-specific interventions. Skill-mix innovation plays a major role in these care models. Three broad trends can be identified. They support the delivery of (i) integrated, multiprofessional care, (ii) improve collaboration of care outside integrated care paradigms and (iii) facilitate continuity and co-ordination of care across organizational boundaries through holistic care approaches.

First, the EU-funded ICARE4EU project identified 101 integrated multimorbidity care initiatives in 25 European countries. Of these, a quarter reported to be well-established care programmes and about 40% to be routine care. Skill-mix innovations were frequently reported among their main objectives, mostly improving multiprofessional collaboration (80%) and care coordination (71%). GPs were involved in most programmes (81%), followed by medical specialists (66%) and nurses (50%). Other occupations included physiotherapists (45%), social workers (40%), home help (37%) and psychologists (33%). Pharmacists and informal carers, who are considered important providers of multimorbidity care, were involved in only one third of the programmes (van der Reference van der Heide, Snoeijs and MelchiorreHeide et al., 2015). In about half of the cases, integrated multimorbidity care initiatives were used to adapt or drive new multiprofessional care delivery structures. Case managers were introduced in about 40% of the initiatives; changes in job descriptions were introduced in a quarter of these initiatives (Reference van der Heide, Snoeijs and Melchiorrevan der Heide et al., 2015).

New teamwork and collaborative care paradigms have also developed next to integrated care networks and disease management programmes with the goal of establishing closer working relationships between primary care (family doctors, GPs and practice nurses) and specialists. Primary care professionals increasingly work together in multiprofessional teams, in complex care hubs/networks and in new patient-centred care models. Most commonly, collaborative care models have been integrated into routine care for the treatment of mental health conditions. In ageing societies, mental illness is frequently comorbid with physical illness, requiring mental health specialists and primary care workers to collaborate closely. Several countries have explored such collaborative care models targeting the reduction in patient hospitalizations. For example, the United Kingdom, Ireland, Italy, the Netherlands, Denmark and Belgium have policies prioritizing community-based, multiprofessional care (Reference Vitale, Mannix-McNamara and CullinanVitale, Mannix-McNamara & Cullinan, 2015). These care delivery approaches provide mental health and medical care in primary care settings at the community level, and coordinate with social care (see Case study 3). They have also been shown to be more effective than usual care for improving depression outcomes (see Section on Skill-mix interventions delivered by multiprofessional teams; Reference Sighinolfi, Nespeca and MenchettiSighinolfi et al., 2014).

Multidisciplinary primary care teams have been set up in various countries (Belgium, Estonia, Finland, France, Italy, the Netherlands, Portugal, Spain and Sweden) to overcome the limitations of solo practices and benefit from interprofessional teamwork and multiprofessional competencies. The principal aim of multidisciplinary primary care teams is to offer a comprehensive set of services and to respond to the specific needs of patients with chronic and multimorbid conditions. There is large variation in team composition across countries; teams may include different social and health professionals besides family physicians and registered nurses (for example, psychologists, health promotors, nutritionists, clinical community pharmacists, physical activity counsellors, community health workers and front desk staff). In particular, the team-based model allows the provision of more complex services such as prevention and health promotion, patient education and self-management support, patient and family caregiver empowerment, psychological counselling, social services, referral and care coordination, which can usually not be offered by single-handed practices (Reference De Maesseneer, Borgermans, Beran, Jakab, Farrington, Borgermans and MantinghDe Maesseneer et al., 2018).

New care coordinating roles to improve the care trajectory of patients with chronic and multimorbid conditions

Reforms and programmes have introduced new professional roles working across organizational boundaries meant to provide coordinated, continuous and holistic care. These new roles range from case managers and social workers to care navigators, coordinators and community facilitators (Reference GilburtGilburt, 2016). Adapted to local health needs, they are especially useful to patients with long-term, multimorbid health conditions as well as people with low levels of literacy or cognitive deficits by helping navigate the fragmented sectors and services involved in covering their care needs (health, social care, housing, employment and education). Their main role is to support patients to plan, organize and access support by linking and coordinating resources and providers through a person-centred approach and provide social and psychological support. Case managers, for example, may also perform clinical needs assessment and clinical monitoring, provide patient and caregiver education and self-management support and establish care goals with the patients coupled with the development of their self-efficacy skills (Reference De Maesseneer, Borgermans, Beran, Jakab, Farrington, Borgermans and MantinghDe Maesseneer et al., 2018).

With this broader perspective, these new roles can also facilitate the implementation of effective, integrated care packages (Reference Ivbijaro, Enum and KhanIvbijaro et al., 2014). The nurse-delivered case management programme in Valencia (Spain) describes an integrated care model that introduced case managers to support patients with multiple chronic conditions, including through arranging care transitions and settings (Case study 4). While these coordinating roles are an emerging trend in many countries, evidence and experiences are still limited.