You are here

  • Home
  • Archive
  • Volume 14, Issue 3
  • Importance of kidney function, number of chronic conditions and medications for hospitalisation in elderly in Blekinge County, Sweden: a case–control study

Article Text

Article menu
Geriatric medicine
Original research
Importance of kidney function, number of chronic conditions and medications for hospitalisation in elderly in Blekinge County, Sweden: a case–control study
  1. Isabell Norstedt1,2,
  2. Kristine Thorell1,
  3. Anders Halling1,2
  1. 1Department of Clinical Sciences Malmö, Center for Primary Health Care Research, Lund University, Malmo, Sweden
  2. 2University Clinic Primary Care Skåne, Region Skåne, Sweden
  1. Correspondence to Dr Anders Halling; anders.halling{at}med.lu.se

Abstract

Objectives To study the association between risk for hospitalisation in an elderly population related to renal function, number of chronic diseases and number of prescribed drugs.

Design A case–control study. Persons hospitalised were included and their controls were obtained from electronic hospital medical records. If data were lacking on creatinine levels, multiple imputation was used.

Setting Blekinge County in southwestern Sweden.

Participants Study of individuals aged 75 years or older in 2013. We identified a total of 2,941 patients with a first hospitalisation. Of these, 81 were excluded, 78 due to incomplete data and 3 because of lack of control persons. Controls were matched to the same sex and birth year, which resulted in 5720 persons.

Primary and secondary outcome measures To analyse the OR for hospitalisation conditional logistic regression was used.

Results A total of 695 persons lacked creatinine value. Using imputation values comparing persons with estimated glomerular filtration rate (eGFR) <30 mL/min/1.73 m2 with ≥30 univariate analyses showed an increased OR 2.35 (95% CI 1.83 to 3.03). Adjusted analyses demonstrated an OR of 1.90 (95% CI 1.46 to 2.47). Comparing eGFR<45 mL/min/1.73 m2 against ≥45 univariate analyses showed OR 1.38 (95% CI 1.22 to 1.57). Adjusted analyses OR for the same group were 1.17 (95% CI 1.03 to 1.33). In both models, the OR for five or more chronic conditions and five or more medications showed a statistically increased risk for hospitalisation.

Conclusions There is a need for systems using data collected in routine care to follow elderly patients to minimise avoidable hospitalisations that can cause adverse effects. Renal function, number of chronic conditions and medications are factors that are of significant importance. This study demonstrates the complexity of this patient group.

  • Aged
  • GERIATRIC MEDICINE
  • Health Services for the Aged
  • Adult nephrology
  • Primary Health Care
  • Adverse events

Data availability statement

No data are available.

http://creativecommons.org/licenses/by-nc/4.0/

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.

https://doi-org.ezproxy.u-pec.fr/10.1136/bmjopen-2023-077421

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    STRENGTHS AND LIMITATIONS OF THIS STUDY

    • A total of 23% of individuals included in this study lacked a creatinine level during the study period. This was handled using multiple imputation.

    • In Blekinge County, Sweden, where this study was conducted, patients had only one electronic medical record for both primary and secondary care, which gives a small risk of missing information concerning diagnoses.

    • Information about medicines was collected via Blekinge County’s register on dispensed medicines, but the use of over-the-counter medicines is unknown as well as the actual intake of prescribed medicine.

    • The study was based on data from routine care and shows a model for estimating the risk of hospitalisation in elderly populations that can be used on similar populations or can be further developed.

    Introduction

    Increased life expectancy

    The number of elderly people is increasing globally as well as in Sweden. Between 2000 and 2015, the estimated life expectancy in Sweden increased by 2.5 years. A reduced mortality rate among people over 65 years of age is the primary cause.1 2 New challenges will arise in healthcare due to the increased number of elderly people and their treatment for use of multiple medications—polypharmacy. Elderly people take up great resources in healthcare.3 Since the elderly population is increasing in numbers, as well more complex clinical features with multimorbidity occur, it is necessary to find strategies for evaluation of where in different disease processes the unique patient is located.4 Not only for one disease but to get a full picture of the person and take into consideration several factors in this assessment, that is, to be able to prioritise and meet the individual patient’s need.5

    Correlation of renal function and other factors

    Physiology changes with age and elderly people more often have increased fat percentage and decreased muscular percentage compared with younger individuals.6–8 Several studies have shown that the glomerular filtration rate (GFR) decreases with age, even though there are different data that are more exact and show at which rate renal function decreases. There is a broad agreement in the existing literature that GFR decreases after 50 years of age.9–13 Renal failure is not only correlated with drug clearance, it is also important due to being correlated with other chronic diseases.14–21 In one study, chronic kidney disease (CKD) only occurred in close to 1% of participants without any other comorbidities and also showed that the multimorbidity cluster changes with the level of estimated GFR (eGFR).22 Drug-related causes have been estimated to include 10%–30% of all acute admissions and are over-represented in elderly patients.14 15 23 Older patients can be exposed to side effects in drugs that are lipophilic (stored in fat tissue) and in drugs that mainly perform renal clearance. There is statistical significance for the number of prescribed drugs in patients diagnosed with CKD depending on age; the patients aged 65 years or older were prescribed considerably more pharmacological treatment.24 With this in mind, one can have an insight indicating that CKD is associated with a greater number of other chronic diseases, especially in the elderly, and the usage of several drugs in this population. Studies of predictors of rehospitalisation have shown a correlation between higher healthcare consumption in the preceding year for patients diagnosed with heart failure, chronic obstructive pulmonary disease, COPD and kidney disease.25 The reasons for hospitalisations are multifactorial, for example, in a study one issue was that good relations and regular meetings in primary healthcare can be associated with fewer days of admission and a lower OR for hospitalisation.16 Earlier studies have shown that patients with chronic conditions who have poor continuity in their healthcare are more prone to visit out-of-hours primary healthcare than other patients.17 18 The term used for describing people with increased risk of adverse events, such as death and acute admission to hospital, is frailty.19 For these patients, there is an increased risk of, for example, acute confusion and delirium after an unplanned hospital admission.20 21 With admission to hospital, especially with repeatedly unplanned admissions because of social factors or inadequate care from a primary healthcare provider, one can hypothesise there is an increased risk for decreased quality of life. This study contributes to an understanding concerning risk factors for hospitalisations in the elderly population. A previous study has shown26 that the number of prescribed drugs and chronic diseases is associated with an increased risk of hospitalisation. What this study adds to our previous research is to investigate if renal function is an additional risk factor for hospitalisation in the elderly.

    Aim

    The aim was to study the association between risk for unplanned admission in hospital care in elderly patients related to renal function, number of chronic diseases and number of prescribed drugs. This study is a continuation of the project: Importance of potentially inappropriate medications, number of chronic conditions and medications, for the risk of hospitalisation in elderly in Sweden: a case–control study.26 This study does not look at the importance of potentially inappropriate medications (PIM) for hospitalisations but instead studies the importance of renal function. We did not study other variables that could be linked to renal function or rehospitalisation because our focus in this study was to the study above mentioned.

    Material and method

    Setting

    Blekinge is a small county in Sweden that had a total of 152 315 inhabitants in 2013. Patients included in the study were individuals aged 75 years of age or older who were registered to a primary healthcare centre in 2013. The Swedish healthcare system is based on that inhabitants are registered to a primary healthcare centre. In Sweden, primary healthcare centres are both public and private and both types are included in this study. In Blekinge County, in 2013, a total of 41% (9) of the primary healthcare centres were private.

    Study population

    Patients aged 75 years of age or older in Blekinge County at their first hospitalisation in 2013 to a medical, geriatric, palliative or orthopaedic ward were included and matched to a control with the same sex and birth year. Controls were taken from the population registered to a primary healthcare centre in Blekinge in 2013. The index date was defined as the date for admission to hospital. For statistical analysis, patients were divided into four age groups: 75–79 years, 80–84 years, 85–89 years and 90 years and older.

    Renal function

    We used P-creatinine levels measured during the period 1 January 2012 until the day prior to an acute admission to assess renal function in patients. If there were two values or more a mean P-creatinine level was calculated. Isotope dilution mass spectrometry, IDMS was used for analysing creatinine values in laboratories in Blekinge during the study period.9 To assess renal function as eGFR we used the Lund-Malmoe formula revised, LM-rev equation.27 Evaluation of renal function in controls was achieved by using the same method. Renal function was divided into four groups, group 1 with eGFR≥90 mL/min/1.73 m2, group 2 with eGFR 60–89 mL/min/1.73 m2, group 3 with eGFR 30–59 mL/min/1.73 m2 and group 4 eGFR<30 mL/min/1.73 m2. Group 1 in our study is not equivalent to chronic kidney disease, CDK stage 1. CDK stage 1 is defined as chronic renal disease without influence on GFR. We chose to put CKD stages 4 and 5 together due to the prevalence in the population. Prevalence of CKD stage 4 is 0.2% and stage 5 is 0.1%.28 29

    Number of chronic conditions

    Information on diagnosis in primary healthcare as well as in hospital care during 2011–2012 was collected from an electronic medical record database in the county council in Blekinge. To assess comorbidity and the number of chronic diseases, a method by Calderón-Larrañaga et al was used to analyse (International Statistical Classification of Disease-10 Revision and Related Health Problems) codes to define if it is a chronic condition in elderly individuals.4 To consider if a disease is chronic or not and discuss its relevance in elderly people, key features were determined such as duration, course, reversibility, treatment and consequences. They were then grouped into 60 groups of chronic conditions. The number of chronic conditions was categorised into five intervals, no chronic diseases, one chronic disease, two to four chronic diseases, five to seven chronic diseases and eight or more chronic diseases.

    Use of prescribed drugs

    To estimate the number of used drugs, data were obtained from the county council register on dispensed medicines that received data from the Swedish eHealth Agency. This register contains the same data on prescribed medication (hence not ‘over-the-counter medication’) as the Prescribed Drug Register at the Swedish National Board of Health and Welfare—a Swedish national register.30 Information on age and sex and dispensed prescribed drugs was found but not dose prescription. To estimate the duration of drug exposure when dose text was not available, we used defined daily drugs (DDD). ‘The DDD is the assumed average maintenance dose per day for a drug used for its main indication in adults’.31 We assumed 0.9 DDDs for regularly used medicines based on calculations for regularly used medicines in an elderly population.32 Medication was classified from anatomical therapeutic and chemical systems.31 An individual’s medication list was created by use of the county council register on prescribed dispensed medicine 3 months before the index date, since drugs are prescribed for use for at most 3 months within the high-cost threshold for medicines in Sweden. If medication was collected more than once during this period, it was only counted as one. The number of drugs was categorised into 5 intervals, no medication, 1–4, 5–9, 10–14 and 15 or more drugs.

    Statistical analysis

    Descriptive demographic statistics and analyses were made with cross-tabulations and χ2 test. A p<0.05 was considered to be statistically significant. Missing data were handled by using multiple imputation with chained equations, which created 50 imputed datasets.33 Cases were matched to a control, 1:1. To analyse OR for hospitalisation, conditional logistic regression was used and presented as OR with a 95% CI. Analyses were presented in different models with univariate analyses regarding renal function in different cut-off groups (<30 mL/min/1.73 m2 compared with ≥30 mL/min/1.73 m2 and <45 mL/min/1.73 m2 compared with ≥45 mL/min/1.73 m2). Three multivariate models are presented, two with imputed values for creatinine value, one with eGFR<30 mL/min/1.73 m2 and one with eGFR<45 mL/min/1.73 m2. The last multivariate analysis included data without imputed values. Calculations were made with STATA V.17 (Stata).

    Patient and public involvement

    There was no patient or public involvement in the development or the conducting of this study.

    Results

    A total of 2941 patients met the inclusion criteria. Of these, 78 were excluded due to incomplete records. Three cases were excluded since there were no suitable control subjects. This left 2860 patients, with an equal number of patients in the control group, which summed up to a study population of 5720 people (figure 1). 21 cases and 674 controls lacked a documented creatinine level during 1 January 2012 until the day before hospitalisation.

    Figure 1
    Figure 1

    Flow chart of selection process.

    The distribution between different wards was 87% to a medical ward, 13% to an orthopaedic ward and 0.49% to a geriatric and palliative ward. Within our study population, 3314 (58%) were female. The mean age was 84 years and the distribution between age categories was 75–79 years—738 (27.38%), 80–84 years—831 (29.06%), 85–90 years—711 (24.86%) and ≥90 years—580 (20.28%), presented in table 1. Table 1 also presents the distribution between renal function, number of chronic conditions and number of medications between cases and controls. The table shows the statistical differences between cases and controls regarding renal function, number of chronic conditions and number of prescribed medications.

    View this table:
    Table 1

    Descriptive statistic of study population

    When using imputed values in a model comparing, calculated with conditional logistic regression, the group with lowest eGFR, <30 mL/min/1.73 m2 against renal function ≥30 mL/min/1.73 m2 the univariate analysis showed increased OR, 2.35 with statistical significance, presented in table 2.

    View this table:
    Table 2

    OR for hospitalisation, cut-off eGFR 30 mL/min/1.73 m2

    In multivariate analyses, eGFR<30 mL/min/1.73 m2 OR for hospitalisation showed increased risk, 1.90 (95% CI 1.46 to 2.47). In the multivariate analyses, OR for hospitalisation increased with the number of chronic conditions, where 5 or more chronic conditions were statistically significant (OR for 5–7 chronic conditions was 1.86, 95% CI 1.48 to 2.32 and 8 or more chronic conditions OR was 2.66, 95% CI 2.04 to 3.45). Five or more medications showed an increased risk for hospitalisation with statistical significance. OR for 5–9 medicines was 1.46 (95% CI 1.21 to 1.77), 10–14 medicines the OR was 2.09 (95% CI 1.62 to 2.69) and 15 or more medicines the OR was 3.82 (95% CI 2.28 to 6.39). When comparing eGFR<45 mL/min/1.73 m2 against renal function ≥45 mL/min/1.73 m2, presented in table 3, univariate analyses showed increased risk with eGFR<45 mL/min/1.73 m2, OR 1.38 (95% CI 1.22 to 1.57). In multivariate analyses, the OR for hospitalisation for the same group was 1.17 (95% CI 1.03 to 1.33). Regarding the number of chronic conditions and the number of medications, the OR was similar to values when comparing eGFR<30 mL/min/1.73 m2.

    View this table:
    Table 3

    OR for hospitalisation, eGFR cut-off 45 mL/min/1.73 m2

    When not using imputed values in a multivariate analysis, the result showed that the only significant factor for admittance to hospital was the use of 10 or more medicines, represented in table 4. OR for 10–14 prescribed drugs was 1.43 (95% CI 1.07 to 1.91) and 15 or more drugs it was 2.69 (95% CI 1.56 to 4.64).

    View this table:
    Table 4

    OR for hospitalisation, without imputed values for eGFR

    Discussion

    In this study, statistically significant risk factors for acute admission in the elderly population were eGFR<30 mL/min/1.73 m2 compared with ≥30 mL/min/1.73 m2 and also eGFR<45 mL/min/1.73 m2 compared with ≥45 mL/min/1.73 m2. In both models, five or more chronic conditions and five or more medicines showed a statistically significantly increased risk for hospitalisation. There was a significant difference between cases and controls regarding renal function, number of medicines and number of chronic conditions, where cases had a higher prevalence of both chronic conditions and medications.

    It was an unexpected finding that almost one out of four persons in the population had not had their creatinine level checked within the last year. Of these, 97% were controls, which can be considered healthier since they did not need hospitalisation. It may not have existed as an indication or occasion for control of renal function. Elderly individuals who are in need of medical care should have their creatinine level checked regularly to detect any alteration in kidney function, and to be able to adjust the dosage of specific prescribed drugs.6 7 9 Despite a decrease in GFR in elderly people, there is not expected to be a corresponding increase in creatinine due to age-related reduction in muscular mass.12 34 35 Persons classified in group 1 in the study were not the same as CKD I. This was not seen as a problem because this study investigates kidney function in relation to other factors (number of chronic conditions and number of medications) and hospitalisation and not if the person had kidney disease with a preserved renal function. Earlier studies have pointed out increased mortality at eGFR<40 mL/min/1.73 m2, but not at eGFR 50–59 mL/min/1.73 m2 in elderly patients.36 This study shows an increased risk for hospitalisation in patients with eGFR<30 mL/min/1.73 m2 and <45 mL/min/1.73 m2 compared with those with better eGFR. Distribution of renal function among the study participants, where most were in the range of impaired renal function, was expected since renal function decreases with age.9–13 The study has not focused on acute or chronic renal failure since it was not a part of the aim of this study. The cause of renal failure was not explored, and furthermore, there was no investigation as to whether the renal failure was prerenal, renal or postrenal.

    An advantage of using imputations was the possibility of handling missing data, which is a common occurrence when using data collected from routine healthcare.37 The exclusion of individuals with incomplete datasets can affect study results, which is the case in this study. When excluding persons without a creatinine level the only significant factor for hospitalisation was 10 or more medications. When using multiple imputation renal function, five or more chronic conditions and medications became statistically significant.

    Data concerning the number of chronic diseases were collected during 2011–2012. Since the observation period began in 2013, there was a potential gap where the subjects could have been diagnosed with a chronic disease earlier in 2013 prior to the admission, which was then unaccounted for in this study. During 2013, Blekinge County did not increase the compensation to health institutions for multimorbidity, hence the risk of upcoding was considered low. The chronic diseases were not classified depending on the need for care and costs associated with the disease, in comparison to other models for estimations.38 It was considered that each chronic disease contributed equally to the health status of the subject and was, therefore, more a reflection of the complexity of care rather than the need for care. It is important to remember that the presence and number of chronic diseases alone were not considered a trustworthy indicator of the need for care compared with the elderly patients’ overall health status.39

    We had information about the dispensal of medicine from pharmacies, but no information on medicine intake. If medicine was dispensed but not taken, overestimation of the number of drugs was a risk. If a patient was prescribed a drug and the collected medication for more than 3 months on the same occasion, there was a risk for underclassification of the number of drugs taken. The reason was that those medications would not appear in the lists of medications created in the study due to a time limit the study. This situation points out a real problem in healthcare, that is, uncertainty of information on which medication the patient is using or not.

    In all the multivariate analyses, the OR for hospitalisation decreased when the patient used 1–4 medications (not statistically significant) and increased when using 5 or more drugs. Earlier studies have shown that polypharmacy is associated with hospitalisation.40 41 Drug-related causes have been estimated to entail a part of these acute admissions and occur to a greater extent in the elderly population.14 15 23 One can ask oneself if treatment with several drugs at the same time results in adverse effects or if it is an expression of multimorbidity and a high burden of disease. In an article by Viktil et al, the authors discuss if fear of polypharmacy causes a less than optimal treatment, and thereby worsens healthcare and treatment of multidisease patients.42 It is important to focus on, study and think of the usage of many drugs at the same time, since having many medications can be beneficial to the patient and have disadvantages and side effects. Furthermore, one can discuss if some medications are not necessary, and some may not be evaluated over time and the prescribing continues without further consideration. If this is the situation, some of the medications may lack remaining indication and therefore the benefits may not weigh up for the potential adverse effects. It is worth considering if there are any special medications in the patient group with a high number of medications, or more medications that themselves increase the risk for hospitalisation rather than the total burden of polypharmacy.

    PIMs are defined as medications that have side effects that outweigh the clinical benefit when used by elderly people.43 In a previous article, with the same original study population as in this study, the results showed that PIM was non-significant for hospitalisation, but also that PIM was associated with the number of medications and chronic conditions. Therefore, the authors concluded that the number of medications can be a confounder for specific medications that cause hospitalisation and that the number of medicines can cover also the effect of PIM since PIM is included in the number of medications.26 Other publications have shown that the number of medications is significant for readmission and not PIM itself.41 A study has reflected that the use of PIM in patients with CKD was almost one out of five.44

    More women than men were included in the study, which is not surprising. Women have a longer life expectancy than men1 and since the study only included elderly people it is to be expected that more women than men would be admitted to hospital. Another aspect of admission and hospitalisation can be sex differences in healthcare consumption.45 Age distribution in the study population comprises patients with a higher age than expected. One explanation could be that elderly people, with a higher prevalence of decreased organ functions, are in extended need of medical attention.46 Data from hospitalisation to other wards than medical, geriatric, palliative or orthopaedic wards, for example, surgery wards are not included, which is a limitation that might cause unwanted exclusion.

    Previous studies of geriatric patients concerning chronic diseases have tried to categorise the diseases and monitor the overall health condition of the patient.4 Blekinge County is a small county in Sweden, both in size and population, and has a fairly comprehensible administrate system. This facilitates the data collection from both the public—as well as from the private healthcare system. A strength of the study population is that it combines data from both primary—and secondary care. Data are from routine care, which we also see as a strength. The limitations of this study were the absence of creatinine levels (23%), uncertainties regarding the compliance of patients’ medication and the risk of covariance between the variables. The study does not consider the socioeconomic status, for example, income and level of education in the subjects, which in earlier studies has been identified as determinants of health for morbidity and increased use of medications.47

    Future research should continue the analysis and the identification of elderly patients who need more extensive primary healthcare and to develop the healthcare system for this large patient category. From the results of this study, it is tempting to suggest a system for predicting the risk of hospitalisation using variables that are routinely collected, which could guide primary healthcare staff in prioritising care to avoid deterioration of the patient’s condition.

    If a model was created, one can use it to prioritise patients in primary healthcare and get guidance on how often these patients preferably should see a doctor or nurse in order to minimise stress and other adverse effects that often occur with hospitalisations. Furthermore, this potential risk classification tool could be used for following individuals over time. It is of the greatest importance to see the full picture of every individual patient when meeting elderly patients with multimorbidity and not only follow specific guidelines.48 Further research should focus on how elderly patients are followed up in the best way in primary healthcare regarding the frequency of visits and content of these visits. As a continuation of this project, it would be of interest to investigate whether the patients acutely admitted to hospital had lower numbers of follow-ups in primary healthcare compared with the control group.

    Conclusion

    There is a need for sensitive systems for continuous follow-up of elderly patients to avoid potentially unnecessary deterioration and hospitalisations in elderly populations, which could have adverse effects. The study highlights the complexity and need for further efforts directed towards the growing fraction of elderly patients in our society. This study adds to previous studies regarding that renal function is independently important for risk of hospitalisation in the elderly in addition to a number of chronic diseases and prescribed drugs.

    Data availability statement

    No data are available.

    Ethics statements

    Patient consent for publication

    Ethics approval

    This study involves human participants and was approved by Regional Ethical Review Board in Lund, Sweden (Dnr 2015/712). The subjects were enrolled retrospectively, and patient charts and registers were accessed after approval from the regional ethical committee. Steps were taken to limit the possible ethical implications by publishing information in two local newspapers, 'Sydöstran' and 'Blekinge läns tidning' where the public was given information about the study and contact information to our group if they wished to be excluded. This was done according to the decision made by the regional ethical committee according to Swedish law when informed that consent was deemed not possible to obtain. The study was conducted retrospectively and thus did not influence the care given during the hospitalisation period. The research data were pseudonymised before it was made available for the researcher and informed consent was not deemed necessary.

    Acknowledgments

    We are grateful to Patrick O’Reilly for his advice in proofreading the manuscript.

    References

    1. ↵
      1. Statistikmyndigheten
      . Sveriges befolkningspyramid. Stockholm, 2019. Available: https://www.scb.se/hitta-statistik/sverige-i-siffror/manniskorna-i-sverige/sveriges-befolkningspyramid/
    2. ↵
      1. OECD/European Observatory on Health Systems and Policies
      . Sverige: Landprofil hälsa 2017. State of Health in the EU [Internet]. Brussels: OECD Publishing, 2017. Available: https://www.oecd-ilibrary.org/content/publication/9789264285453-sv
    3. ↵
      1. Socialstyrelsen
      . Kapaciteten i sjukvården socialstyrelsen - att hantera kritiska lägen. Stockholm, 2018. Available: https://www.socialstyrelsen.se/globalassets/sharepoint-dokument/artikelkatalog/ovrigt/2018-12-1.pdf
    4. ↵
      1. Calderón-Larrañaga A,
      2. Vetrano DL,
      3. Onder G, et al
      . Assessing and measuring chronic Multimorbidity in the older population: a proposal for its operationalization. J Gerontol A Biol Sci Med Sci 2017;72:1417–23. doi:10.1093/gerona/glw233
      OpenUrlPubMed
    5. ↵
      1. Muth C,
      2. van den Akker M,
      3. Blom JW, et al
      . The Ariadne principles: how to handle multimorbidity in primary care consultations. BMC Med 2014;12:223. doi:10.1186/s12916-014-0223-1
    6. ↵
      1. Shi S,
      2. Mörike K,
      3. Klotz U
      . The clinical implications of ageing for rational drug therapy. Eur J Clin Pharmacol 2008;64:183–99. doi:10.1007/s00228-007-0422-1
      OpenUrlCrossRefPubMed
    7. ↵
      1. Davies EA,
      2. O’Mahony MS
      . Adverse drug reactions in special populations - the elderly. Br J Clin Pharmacol 2015;80:796–807. doi:10.1111/bcp.12596
      OpenUrlCrossRefPubMed
    8. ↵
      1. Mangoni AA,
      2. Jackson SHD
      . Age-related changes in pharmacokinetics and pharmacodynamics: basic principles and practical applications. Brit J Clinical Pharma 2004;57:6–14. doi:10.1046/j.1365-2125.2003.02007.x
      OpenUrl
    9. ↵
      1. SBU
      . Skattning av njurfunktion. En systematisk litteraturöversikt. Stockholm: Swedish Council on Health Technology Assessment (SBU), 2013.
    10. ↵
      1. Poggio ED,
      2. Rule AD,
      3. Tanchanco R, et al
      . Demographic and clinical characteristics associated with glomerular filtration rates in living kidney donors. Kidney Int 2009;75:1079–87. doi:10.1038/ki.2009.11
      OpenUrlCrossRefPubMedWeb of Science
    11. ↵
      1. Berg UB
      . Differences in decline in GFR with age between males and females. Reference data on clearances of inulin and PAH in potential kidney donors. Nephrol Dial Transplant 2006;21:2577–82. doi:10.1093/ndt/gfl227
      OpenUrlCrossRefPubMedWeb of Science
    12. ↵
      1. Nilsson-Ehle P,
      2. Berggren Söderlund M,
      3. Theodorsson E, et al
      . Laurells Klinisk kemi i praktisk medicin. Lund: Studentlitteratur, 2012.
    13. ↵
      1. Fehrman-Ekholm I,
      2. Seeberger A,
      3. Björk J, et al
      . Serum Cystatin C: a useful marker of kidney function in very old people. Scand J Clin Lab Invest 2009;69:606–11. doi:10.1080/00365510903015771
      OpenUrlCrossRefPubMed
    14. ↵
      1. Pirmohamed M,
      2. James S,
      3. Meakin S, et al
      . Adverse drug reactions as cause of admission to hospital: prospective analysis of 18 820 patients. BMJ 2004;329:15–9. doi:10.1136/bmj.329.7456.15
      OpenUrlAbstract/FREE Full Text
    15. ↵
      1. Paul E,
      2. End-Rodrigues T,
      3. Thylén P, et al
      . Läkemedelsbiverkan Vanlig Orsak till Sjukhusvård AV Äldre. en Klinisk Retrospektiv Studie. Läkartidningen 2008;105:2338–42.
      OpenUrl
    16. ↵
      1. Ranstad K
      . Importance of relationships with primary care. Implications for patients and health care. Lund, Sweden: Lund University, Sweden, 2017.
    17. ↵
      1. Kohnke H,
      2. Zielinski A
      . Association between continuity of care in Swedish primary care and emergency services utilisation: a population-based cross-sectional study. Scand J Prim Health Care 2017;35:113–9. doi:10.1080/02813432.2017.1333303
      OpenUrlCrossRefPubMed
    18. ↵
      1. Flarup L,
      2. Moth G,
      3. Christensen MB, et al
      . Chronic-disease patients and their use of out-of-hours primary health care: a cross-sectional study. BMC Fam Pract 2014;15:114. doi:10.1186/1471-2296-15-114
    19. ↵
      1. Rockwood K,
      2. Mitnitski A,
      3. Song X, et al
      . Long-term risks of death and Institutionalization of elderly people in relation to deficit accumulation at age 70. J Am Geriatr Soc 2006;54:975–9. doi:10.1111/j.1532-5415.2006.00738.x
      OpenUrlCrossRefPubMedWeb of Science
    20. ↵
      1. Swedish Council on Health Technology Assessment (SBU)
      . Omhändertagande av äldre som inkommer akut till sjukhus - med fokus på sköra äldre. En systematisk litteraturöversikt. Stockholm, 2013. Available: https://www.sbu.se/contentassets/5f0e7213e73b4369acd4874fd3dcbf89/akutvard_aldre.pdf
    21. ↵
      1. Villacampa-Fernández P,
      2. Navarro-Pardo E,
      3. Tarín JJ, et al
      . Frailty and multimorbidity: two related yet different concepts. Maturitas 2017;95:31–5. doi:10.1016/j.maturitas.2016.10.008
      OpenUrlPubMed
    22. ↵
      1. Corsonello A,
      2. Fabbietti P,
      3. Formiga F, et al
      . Chronic kidney disease in the context of multimorbidity patterns: the role of physical performance: the screening for CKD among older people across Europe (SCOPE) study. BMC Geriatr 2020;20:350. doi:10.1186/s12877-020-01696-4
    23. ↵
      1. Odar-Cederlöf I,
      2. Oskarsson P,
      3. Öhlén G, et al
      . Läkemedelsbiverkan Som Orsak till Inläggning På Sjukhus. Vanliga Medel Står För Merparten, Visar Tvärsnittsstudie. Läkartidningen 2008;105:890–3.
      OpenUrl
    24. ↵
      1. Formica M,
      2. Politano P,
      3. Marazzi F, et al
      . Acute kidney injury and chronic kidney disease in the elderly and polypharmacy. Blood Purif 2018;46:332–6. doi:10.1159/000492149
      OpenUrlPubMed
    25. ↵
      1. Stäck P,
      2. Forsberg B,
      3. Högberg M, et al
      . Risken För Akut Återinläggning Kan Förutsägas. Tidigare Vårdkonsumtionsmönster och Vissa Diagnoser Starkt Predicerade. Läkartidningen 2012;109:2211–5.
      OpenUrl
    26. ↵
      1. Thorell K,
      2. Midlöv P,
      3. Fastbom J, et al
      . Importance of potentially inappropriate medications, number of chronic conditions and medications for the risk of hospitalisation in elderly in Sweden: a case-control study. BMJ Open 2019;9:e029477. doi:10.1136/bmjopen-2019-029477
    27. ↵
      1. Björk J,
      2. Grubb A,
      3. Sterner G, et al
      . Revised equations for estimating glomerular filtration rate based on the Lund-Malmo study cohort. Scand J Clin Lab Invest 2011;71:232–9. doi:10.3109/00365513.2011.557086
      OpenUrlCrossRefPubMed
    28. ↵
      1. Tidig upptäckt och initial utredning av kronisk njursjukdom
      . Riktlinjer För Preventiv Nefrologi I Södra Sjukvårdsregionen. 2011. Available: http://www.egfr.se/tidig_uppt_initial_utredn_njursjd_sept2016.pdf
    29. ↵
      1. Levey AS,
      2. Coresh J,
      3. Balk E, et al
      . National kidney foundation practice guidelines for kidney disease: evaluation, classification, and stratification. Ann Intern Med 2003;139:137–47. doi:10.7326/0003-4819-139-2-200307150-00013
      OpenUrlCrossRefPubMedWeb of Science
    30. ↵
      1. Wettermark B,
      2. Hammar N,
      3. Fored CM, et al
      . The new Swedish prescribed drug register--opportunities for pharmacoepidemiological research and experience from the first six months. Pharmacoepidemiol Drug Saf 2007;16:726–35. doi:10.1002/pds.1294
      OpenUrlCrossRefPubMedWeb of Science
    31. ↵
      1. WHO Collaborating Centre for Drug Statistics Methodology
      . Guidelines for ACT classification and DDD assignment 2020. Oslo, Norway, 2019. Available: https://www.whocc.no/filearchive/publications/2020_guidelines_web.pdf
    32. ↵
      1. Johnell K,
      2. Fastbom J
      . The use of benzodiazepines and related drugs amongst older people in Sweden: associated factors and concomitant use of other Psychotropics. Int J Geriatr Psychiatry 2009;24:731–8. doi:10.1002/gps.2189
      OpenUrlCrossRefPubMed
    33. ↵
      1. Austin PC,
      2. White IR,
      3. Lee DS, et al
      . Missing data in clinical research: a tutorial on multiple imputation. Can J Cardiol 2021;37:1322–31. doi:10.1016/j.cjca.2020.11.010
      OpenUrlCrossRefPubMed
    34. ↵
      1. Simonsson P
      . LäR KäNna din patients GFR! Läkartidningen 2009;106:425–7.
      OpenUrl
    35. ↵
      1. Perrone RD,
      2. Madias NE,
      3. Levey AS
      . Serum creatinine as an index of renal function: new insights into old concepts. Clin Chem 1992;38:1933–53.
      OpenUrlAbstract/FREE Full Text
    36. ↵
      1. O’Hare AM,
      2. Bertenthal D,
      3. Covinsky KE, et al
      . Mortality risk stratification in chronic kidney disease: one size for all ages J Am Soc Nephrol 2006;17:846–53. doi:10.1681/ASN.2005090986
      OpenUrlAbstract/FREE Full Text
    37. ↵
      1. Wells BJ,
      2. Nowacki AS,
      3. Chagin K, et al
      . Strategies for handling missing data in electronic health record derived data. eGEMs 2013;1:7. doi:10.13063/2327-9214.1035
      OpenUrl
    38. ↵
      1. Starfield B,
      2. Weiner J,
      3. Mumford L, et al
      . Ambulatory care groups: a categorization of diagnoses for research and management. Health Serv Res 1991;26:53–74.
      OpenUrlPubMedWeb of Science
    39. ↵
      1. Young Y,
      2. Frick KD,
      3. Phelan EA
      . Can successful aging and chronic illness coexist in the same individual? A multidimensional concept of successful aging. J Am Med Dir Assoc 2009;10:87–92. doi:10.1016/j.jamda.2008.11.003
      OpenUrlCrossRefPubMedWeb of Science
    40. ↵
      1. Flaherty JH,
      2. Perry HM,
      3. Lynchard GS, et al
      . Polypharmacy and hospitalization among older home care patients. J Gerontol A Biol Sci Med Sci 2000;55:M554–9. doi:10.1093/gerona/55.10.m554
      OpenUrlCrossRefPubMedWeb of Science
    41. ↵
      1. Fabbietti P,
      2. Di Stefano G,
      3. Moresi R, et al
      . Impact of potentially inappropriate medications and polypharmacy on 3-month readmission among older patients discharged from acute care hospital: a prospective study. Aging Clin Exp Res 2018;30:977–84. doi:10.1007/s40520-017-0856-y
      OpenUrl
    42. ↵
      1. Viktil KK,
      2. Blix HS,
      3. Moger TA, et al
      . Polypharmacy as commonly defined is an indicator of limited value in the assessment of drug-related problems. Br J Clin Pharmacol 2007;63:187–95. doi:10.1111/j.1365-2125.2006.02744.x
      OpenUrlCrossRefPubMedWeb of Science
    43. ↵
      1. Socialstyrelsen
      . Indikatorer för god läkemedelsterapi hos äldre. Stockholm, 2017. Available: https://www.socialstyrelsen.se/globalassets/sharepoint-dokument/artikelkatalog/ovrigt/2017-6-7.pdf
    44. ↵
      1. Aucella F,
      2. Corsonello A,
      3. Soraci L, et al
      . A focus on CKD reporting and inappropriate prescribing among older patients discharged from geriatric and Nephrology units throughout Italy: a nationwide multicenter retrospective cross-sectional study. Front Pharmacol 2022;13:996042. doi:10.3389/fphar.2022.996042
    45. ↵
      1. Osika Friberg I,
      2. Krantz G,
      3. Määttä S, et al
      . Sex differences in health care consumption in Sweden: a register-based cross-sectional study. Scand J Public Health 2016;44:264–73. doi:10.1177/1403494815618843
      OpenUrlCrossRefPubMed
    46. ↵
      1. Dehlin O,
      2. Rundgren Å
      . Geriatrik. Lund: Studentlitteratur, 2007.
    47. ↵
      1. Thorell K,
      2. Skoog J,
      3. Zielinski A, et al
      . Licit prescription drug use in a Swedish population according to age, gender and socioeconomic status after adjusting for level of multi-morbidity. BMC Public Health 2012;12:575. doi:10.1186/1471-2458-12-575
    48. ↵
      1. Eklund-Grönberg A,
      2. Köhler M,
      3. Ekdahl A, et al
      . Bättre Vård För Multisjuka Äldre. Läkartidningen 2009;106:2512–3.
      OpenUrl

    Supplementary materials

    • Supplementary Data

      This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Footnotes

    • Contributors In accordance with Vancouver Protocol, IN, AH and KT all have contributed to the design of this study. IN, AH and KT researched data. AH conducted the statistical analyses and IN and AH wrote the manuscript. AH is guarantor.

    • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

    • Competing interests None declared.

    • Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.