The Co-Occurrence of Polypharmacy and Unmet Needs for Social Care in Older People: A Systematic Review

Polypharmacy is common in older people who often live with disability and dependency. Te number of older people living with unmet needs for social care is also believed to be rising. Polypharmacy is simple to operationalise, whilst unmet needs are not routinely identifed but are known to adversely afect health and well-being. Terefore, this systematic review aimed to investigate whether polypharmacy is a marker of unmet needs for social care in older people. Sixteen databases were searched from inception to January 2021. Studies were included if they reported quantitative data for polypharmacy (“multiple medicines”) in relation to unmet needs for social care (“relative or absolute”) in older people (“study criteria aged ≥ 55years or mean age ≥ 55 in the sample as a whole or stratifed data for the ≥ 55-year age group”) and were from a high-income country (defned by the World Bank). Quality was assessed using the National Institute for Health tool for observational studies. Four studies were identifed from 2,549 citations, and overall, the quality of evidence was low. Some older people using multiple medications had their social care needs met, whilst others did not. However, there is a clear rationale as to why polypharmacy may be linked to unmet social care needs. Given the limited studies identifed in this review, future research should explore this further. Te type of unmet need measure may be important to understand the nature of the relationship between the use of multiple medications and unmet social care needs.

Te reasons underlying the unmet need for social care are varied.Te availability of care is one such reason.For example, England's social care crisis is well-documented, with state-funded services preserved for those with the most severe needs and the fewest assets [18].Other high-income countries are facing similar challenges [19].Other factors that underpin unmet needs include the accessibility of services and attitudes or expectations about the acceptability of needing or requesting services [20].Older people with multiple health problems may prefer to prioritise those issues that are most important to them, leaving some problems unaddressed.Te supply of unpaid care is also important in understanding the rise of unmet social care needs.Older spouse carers are likely to be living with disabilities themselves [3], whilst the supply of unpaid care by adult children cannot keep pace with the growing need for care [21].Tis is a pattern that is expected to worsen: for example, as dependency continues to rise with population aging [22], projections suggest a decline in the availability of formal and informal caregivers [21][22][23][24].

Polypharmacy: A Clinical Marker of Unmet Social Care
Needs? Unmet needs are an important indicator of equity in care but are not routinely identifed in practice [25].Polypharmacy by contrast is simple to operationalise and is more commonly integrated into care records.Critically, polypharmacy has a rationale link to unmet social care needs and could potentially serve as a useful clinical marker of such needs.
Te most compelling argument for a link between polypharmacy and unmet needs is that multiple medicines will likely indicate multiple health conditions and a range of associated needs.We know that poorer health and a greater number of functional difculties are linked to a greater risk of unmet social care needs [26].Terefore, it would be reasonable to assume that polypharmacy-a marker of poorer health and more functional limitations-could also be linked to an unmet need for care support.Tis makes sense, especially in the context of stretched and fragmented health and social care services [27], where someone with multiple care needs (and competing priorities) may have only a proportion of those needs met.
We also know that some people struggle with regimens of taking multiple medications [28].When adherence is compromised, we can expect to see some adverse impact on health and the ability to live well day-to-day.Such difculties in managing multiple medications may, therefore, impact on health and increase the need for support that is not necessarily met through services or informal care.
Furthermore, where certain combinations of medicines have undesirable consequences, people using multiple medications may adjust regimes on their own terms, accepting a compromise to day-to-day functioning and well-being.Polypharmacy may, therefore, signal the potential for unmet care needs, but where such (unmet) care needs are an acceptable compromise.
Finally, we know that polypharmacy is not always managed well from the service perspective [29]: if multiple medications are inadequately or infrequently reviewed, this may compromise the person's health and well-being, resulting in needs that could go unmet.

Why Is It Important to Investigate the Co-Occurrence of Polypharmacy and Unmet Social Care Needs in Older
Individuals?As outlined above, there is a rational link between polypharmacy and unmet social care needs, which may have a useful clinical application.Identifying populations with unmet needs can help to minimise the adverse health consequences when people do not receive timely help with day-to-day activities.However, older populations are not routinely screened for unmet social care needs, thus missing important opportunities for intervention.In contrast, data about polypharmacy are widely collected and accessible across care settings.Tus, if polypharmacy is clearly linked to unmet social care needs, it could serve as a useful clinical marker to identify at risk populations.To consider this hypothesis, this systematic review aimed to synthesise evidence about the association and co-occurrence of polypharmacy and unmet social care needs in older people.

Materials and Methods
2.1.Protocol.Te protocol for this review was registered with PROSPERO (CRD42021230606).

Review Criteria.
Studies in any language were included if they reported quantitative data for polypharmacy ("the use of multiple medicines") [30] in relation to unmet needs for social care ("relative or absolute") [8] in older people ("study criteria aged ≥55 years or mean age ≥55 in the sample as a whole or stratifed data for the ≥55-year age group") and were from a high-income country as defned by the World Bank [31]."Relative unmet needs" describe the situation where the help received is judged to be insufcient, whilst "absolute unmet needs" describe the situation where help is needed but not received.Receipt of help included both paid and unpaid support.Our decision to use the lower threshold of 55+ years acknowledges that the onset of disability from long-term conditions starts earlier in the life course for the most socioeconomically disadvantaged populations [32].Te younger age threshold of 55+ therefore minimises the risk of eliminating evidence from such disadvantaged groups.No time limits were applied other than those imposed by the limits of the databases.Randomised controlled trials, qualitative studies, news items, editorials, opinion pieces, and irretrievable full texts were excluded.No other restrictions were placed on the study designs eligible for inclusion.

Search Strategy.
To identify academic publications and grey literature, subject headings and keywords for (i) polypharmacy, (ii) unmet needs and (iii) older people were combined, using tailored strategies developed with and translated by an information scientist (Appendix 1).Sixteen databases were searched from inception to January 2021.Tese included Medline, Embase, PsycInfo, CINHAL, Scopus, Cochrane Database of Systematic Reviews (CDSR), Health Management Information Consortium (HMIC), Social Care Online, Social Care Institute for Excellence (SCIE), NHS Evidence, Health Survey for England, Te Health Foundation, Te King's Fund Library, OpenGrey, Te British Library electronic theses database (ETHOS), and Google Scholar (frst 300 results) [33].Reference lists of included studies and publications of authors known to have carried out work on this topic were hand-searched to identify further potential publications.A fnal search in February 2022 to update the review identifed no further eligible studies.

Study Selection. Titles and abstracts of all search records
were screened by one reviewer (anonymised), and a sample of 50% was screened by a second reviewer (anonymised) to check consensus.Te full texts of potentially relevant papers were then examined independently by both reviewers, and disagreements were resolved through discussion.Records not published in English were translated using Google Translate as necessary.[35].No records were excluded on the basis of quality assessment in order to present the evidence in context.

Evidence Synthesis.
A meta-analysis was inappropriate due to the limited reporting of an efect size.A narrative synthesis was, therefore, undertaken [36].

Results
From 2549 citations, four studies conducted in the community setting, and published between 2013 and 2019, were included (Figure 1).All of the included studies operationalised polypharmacy as medication count, and one examined medication classes within this [37].Tree studies were reported as cross-sectional [38][39][40], whilst one [37] did not explicitly report the design; this was interpreted as crosssectional by the review team.
One study, which used an absolute measure of unmet need, indicated that polypharmacy was similar between those with and without unmet care needs [40].In two studies, the populations were typically taking nine or more medicines daily, yet the proportion reporting unmet needs was 11% in one study [37] and up to 90% in the other [39].Te study by Naess and colleagues used a relative measure of unmet needs, which we know is shaped by expectations and judgements of the adequacy of care.Te study by Kayyali and colleagues appeared to use an absolute measure of unmet need, which ofers a more objective quantifcation of unmet need (although it did not explicitly state the type of measure used).Te diferences in rates of unmet need Health & Social Care in the Community Health & Social Care in the Community among these two study populations with similar levels of polypharmacy may, therefore, be partly explained by the potentially diferent unmet need measures used.Te study by Jamieson and colleagues [38] operationalised unmet need as a relative measure in only one IADL domain (managing medications), unlike the other studies where measures of unmet need included multiple IADLs.Tis study confrms that more medications are associated with needing more help to manage them.Two studies were rated fair in quality [38,40], and two were rated poor [37,39].Characteristics of the included studies are summarised in Table 1, and a quality assessment table can be found in Appendix 2.

Principle Findings.
Tere is limited evidence about the link between polypharmacy and unmet social care needs from the few studies included in this review.Tere is a rational link between unmet social care needs and polypharmacy which would warrant future research, but a very small, poor quality, and heterogeneous evidence base limits our assessment of this.

Why Might Polypharmacy Potentially be Linked to
Unmet Social Care Needs?As to why some older people in each included study of this review experienced unmet needs for social care, those prescribed multiple medicines may, for example, have complex conditions and therefore competing priorities, particularly as health and social care are too often fragmented [27].Furthermore, the evidence included in this review came from the community setting, and most older people (who often live with polypharmacy) remain in their own homes as they age [41], though not all may have suffcient access to formal or informal caregiver support.
Social care might also be a component of what is driving polypharmacy.Unmet social care needs could lead to deteriorating health [5,6,[9][10][11][12] and then increased healthcare utilisation [13][14][15]17] and additional prescriptions, for example.Tat said we cannot determine from the limited evidence identifed in this review whether a person would have fewer unmet needs if they took fewer medicines.Te need for more help with respect to practical aspects of medicine taking among community-dwelling stroke survivors [38] may refect the wider literature in terms of the informal caregivers' burden and lack of training and support [42].Meanwhile, among those prescribed multiple medicines where mobility problems impaired self-care [37], it is possible that anticholinergic medications, for example, contributed to functional impairment [43,44].

Why Might the Evidence be Mixed in Tis Review?
Contrasting fndings about rates of unmet need in populations with similar levels of polypharmacy within this review [37,39] may be, speculatively, explained by the measure of unmet need.However, more evidence is needed to clarify this.Using (what appears to be) an absolute approach to operationalise unmet needs [37] would, for example, capture people who are most in need of support [8], unlike a relative measure [39], which is driven by expectations of care.Furthermore, whilst, in a nationally representative sample of older people, the levels of polypharmacy were similar in those with and without absolute unmet needs, polypharmacy was markedly lower among those who did not need help with ADLs [40].Tus, we can infer that polypharmacy is indicative of high needs, but these are potentially not always assessed and met.

Strengths and Limitations.
We employed a comprehensive search strategy with clear review criteria to ensure all relevant evidence was represented.Te evidence we found came from the community setting, which is of interest given that "aging in place" is the preferred strategy in many countries [45].Given the projected rise in social care needs with population aging [3,46], our examination of unmet needs for social care as a standalone outcome, and exclusion of studies where the social care component could not be isolated (for example, see [47,48]) is another strength.
A limitation of this review is the paucity and limited quality of evidence, due to the cross-sectional nature of the included studies for example, which meant we could not assess causation or the duration of unmet need (Appendix 2).It is also possible that the evidence presented is understated, as the included studies were published from 2013 to 2019, and unmet needs are believed to have risen during the COVID-19 pandemic [49,50].From the evidence included, we could not determine a specifc medication threshold most associated with unmet social care needs, though an unmet need for support to manage multiple medications appears more likely for those taking more medications [38].Nor could we draw any conclusions about the relationship between polypharmacy and unmet needs among older people living in care homes-a group often excluded from research studies.Home nursing care [39] is also not exclusively limited to personal assistance.Finally, in the included studies, it was not always clear whether unmet needs stemmed from insufcient formal or informal care provision [38], which perhaps refects the blurring of boundaries between the two [51].

Conclusion
In this review, some older people using multiple medications had their social care needs met, whilst others did not.However, there is a clear rationale as to why polypharmacy may be linked to unmet social care needs.Given the limited studies identifed in this review, future research should explore this further, for example with population-level longitudinal datasets that seek to identify (i) whether unmet needs are more common beyond a certain medication threshold, and if so, (ii) whether this difers between settings and (iii) the pre and post COVID-19 era.Te type of unmet need measure may also be important to understand the nature of the relationship between the use of multiple medications and unmet social care needs.Future studies should explore this where data are available.

6
Health & Social Care in the Community

Table 1 :
Characteristics of included studies.