Risk factors for the development of delirium in elderly patients undergoing orthopaedic surgery: A systematic review and meta-analysis

1. Introduction

Postoperative delirium (POD) is defined as an acute neuropsychiatric syndrome characterized by a temporary and reversible functional brain disorder that manifests with sudden disturbances in consciousness and orientation resulting from physical or physiopathological factors after surgical interventions [1-3]. Although the physiopathology of delirium has not been fully clarified as yet, several hypotheses have been proposed. Of these hypotheses, disorders of cerebral oxidative metabolic processes and aberrations involving more than one neurotransmitter are the major ones [4-7]. The risk factors leading to delirium are examined in two separate groups. These groups are the predisposing risk factors and the risk factors that trigger delirium. The predisposing risk factors increase the tendency to develop delirium. Already existing at the time of hospital admission and acting on the susceptibility of the patient, these factors include male gender, fractures, depression, visual disturbances, dementia, polypharmacy, and age as the major ones. The risk factors that trigger delirium are those that trigger, facilitate, and lead to the development of delirium. Of these factors, surgical interventions, fluid-electrolyte disturbances, pain, invasive catheterization (CVP, indwelling urinary catheters, nasogastric intubation, etc.), medications (especially neuroleptics, narcotics, etc.); the use of hearing devices, eye-glasses, etc.; and the length of stay in the inpatient unit or the intensive care unit are some of the major ones [2,3].

Delirium is seen mostly in surgical clinics after intensive care units in hospitalized patients [6,7]. In the literature, it has been reported that the prevalence of delirium is 50% in hospitalized geriatric patients, and that 3-53% of patients undergoing major surgery and 83% of critical patients in intensive care are impacted by delirium [8,9]. Although many factors cause delirium, delirium is observed more commonly in some patient groups, primarily geriatric patients and patients undergoing cardiac and hip surgery [1,10]. Delirium can be seen in any age group in an age-dependent pattern. It has been reported in the literature that the incidence increases at the age of 60 years and older, and that 80 years of age is an important risk factor [5]. The relationship between advanced age and delirium is associated with the physiological and cognitive changes which occur with ageing. During the ageing process, changes also occur in the brain, similar to physiological alterations occurring in the heart, kidneys, the liver, and other organs. There is known to be a decline in the brain weight, the blood supply to the brain, the number of neurons, and the neurotransmitter quantities together with ageing [10,11]. Delirium has a negative effect not only on patients, but also on their relatives, and the healthcare system in emotional, functional, and financial aspects. These effects lead to several consequences including longer hospital stay, and increased mortality, morbidity, and hospital costs [12-14]. Cognitive and functional losses, which are the long-term effects of delirium, may continue further after the hospital discharge leading to rehospitalization [13,14]. Affecting wellbeing unfavorably, the clinical symptoms and risk factors of delirium have not yet been fully recognized and are often not identified by healthcare professionals as they can be misdiagnosed as other types of cognitive disorders [5, 15]. The aim of this study was to systematically review the national and international studies that have investigated the risk factors leading to delirium in geriatric patients after orthopaedic surgery and to perform a meta-analysis using the data reported by those studies. It can be considered that systematic reviews and meta-analyses for the investigation of risk factors of delirium will contribute to the literature, increase the awareness of healthcare professionals, and pave the way for future studies.

2. Materials and Methods

The type of the study, the study site, and the duration of the study This study used the meta-analysis method, which is a quantitative study method. The study was carried out in the period from October 2019 to December 2019.

Ethical Requirements The literature review model was used as this study is a meta-analysis study. As the literature review does not affect animals or humans directly, ethics committee approval was not sought.

Steps of the Study The steps of the study were categorized based on the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement) and MOOSE (Meta-analysis of Observational Studies in Epidemiology) criteria concerning the papers to be included in the study. The articles meeting those criteria were identified and are presented in Figure1 and Table 1.

Detailed Literature Review Using the selected keywords in English 'Orthopaedic Surgery, Geriatrics, and Delirium'; the literature review was performed on all of the planned databases. Relevant research articles published in the period from January 2010 to November 2019 were retrieved from CINAHL, ScienceDirect, Google Academics, Web of Science, PubMed, and Ovid and were included in the assessment process. After retrieving the relevant articles from each of these websites, any repetitions were excluded, before proceeding to the step of reading the article titles, abstracts, and full-text manuscripts [16-19].

A total of 346 articles were identified with the planned search strategy on the six databases. After excluding the repeated papers and those with non-relevant titles and abstracts, the remaining 282 papers were included in the assessment. The titles and abstracts were read to select the papers to be included in the full text-reading process. The articles with a non-relevant subject matter were categorized in detail and excluded from the study. The articles search and the elimination processes used in the selection of the papers to be included in the study are presented in Figure 1 [16,20,21].

Study title Study Objective Assessment tools used in the study Sample characteristics Study outcomes Approach to The Risk Of Delirium In An Orthogeriatric Unit Bert et al. 2019 [22]

The aim of this study was to determine the efficacious method between two different methods used for the management of delirium risk in an orthogeriatric unit.

The Confusion Assessment Method Scale Orthogeriatric patients (N: 85). Cohort study

The experimental treatment was successful as the incidence of delirium was reduced and the reduction in the delirium duration was found to be statistically significant.

Relationship of Postoperative Delirium And Cognitive Impairment And Their Impact On Functional Status In Older Patients Undergoing Orthopaedic Surgery: A Prospective Cohort Study Liang et al. 2014 [19]

The aim of this study was to evaluate the post-discharge functioning of the patients developing delirium after orthopaedic surgery and to assess the relationship between postoperative delirium and underlying cognitive disorders.

Chinese Geriatric Depression Scale Activities of Daily Living Confusion Assessment Method Mini- Mental State Examination ASA Lawton-Brody instrumental ADL (IADL) Old patients (N: 232) undergoing orthopaedic surgery. Cohort study

Early assessment of cognitive functional status is needed to improve the postoperative functional status of patients and prevent delirium after orthopaedic surgery. Postoperative delirium can be predicted by reductions in IADL scores.

Prevalence of delirium, risk factors, and cognitive functions in elderly hip fracture patients with general and spinal anesthesia Atay et al. 2012 [21]

The aim of this study was to investigate the prevalence of postoperative delirium in Turkey; developing after hip fractures in association with the presence of some risk factors and to evaluate the effects of anesthesia on postoperative cognitive functions.

Structured clinical interview scale based on DSM-IV Standardized Mini-Mental Test (SMMT) Delirium Rating Scale (DRS) Old patients undergoing orthopaedic surgery (N:70). Descriptive study

A history of previous delirium and advanced age are unpreventable risk factors. Supportive treatment for metabolic disturbances in hip fracture patients may reduce the risk of postoperative delirium. However, it was determined that the type of anesthesia was not a risk factor for predisposing to delirium and cognitive impairment in the postoperative period.

Unraveling the Relationship Between Delirium, Brain Damage, and Subsequent Cognitive Decline in a Cohort of Individuals Undergoing Surgery for Hip Fracture. Beishuizen et al. 2017 [20]

To evaluate the relationship between serum S100B levels (a marker of brain injury), delirium, and subsequent cognitive decline.

Informant Questionnaire on Cognitive Decline Short Form Mini-mental state examination S100B levels in serum samples Patients, who were 65 years and older (65- 102 years), and who were admitted for hip fracture surgery (n = 385). Descriptive study

Serum S100B levels were not correlated with developing delirium in a cohort of old patients with hip fractures. S100B was associated with cognitive decline or death in the first year after hip fracture only in participants without perioperative delirium. S100B appears to have limited value as a biomarker of delirium-related brain injury.

Comparison of incidence and risk factors of delirium between general and regional anesthesia in elderly patients after lower extremity surgery Kupeli et al. 2016 [5]

The aim of this study was to compare the risk factors and delirium incidence in the postoperative period in old patients undergoing orthopaedic

1.1. surgery

under general or regional anesthesia.

Confusion Assessment Method Delirium Rating Scale-Revised-98 Old-age patients (N:120) at the age of ≥65 years; for whom total hip or knee arthroplasty and femoral fracture surgery were scheduled. Descriptive study

We found that the incidence of delirium was 12.5% in old-age patients, who underwent total hip or knee arthroplasty and femoral fracture

2.1. surgery.

Advanced age and polypharmacy are the risk factors for delirium. Delirium increases the length of hospital stays and costs. Study Search and Inclusion criteria for the Meta-analysis i. Studies presented in original papers, ii. Papers in English or Turkish iii. Relevant papers published in national/international peer-reviewed journals accessible by the university as an institution, iv. Papers of studies which included patients aged 65 years and older who were diagnosed with delirium after orthopaedic surgery, v. Papers presented over the last 9 years (2010- 2019).

vi. The full-text papers reporting the total participant numbers in the groups and reporting the changes in the predetermined study outcomes completely and quantitatively were included in the present study.

vii. Patients with no history of delirium or dementia before the surgical intervention, viii. Patients with no history of neurological or mental disorders,

The Review, Coding, and the Methodological Quality Assessment of the Papers as per the Inclusion Criteria Independent and detailed readings of abstracts and full-text manuscripts of the papers were completed by two researchers/specialists to prevent publication bias. The papers included in the assessment were coded according to descriptive characteristics, as follow: i. The title and the date of the study ii. The type and the objective of the study iii. The sample size of the study iv. Whether the study was interventional v. The scales used in the study vi. The primary outcome of the study vii. The quality assessment score After examining the selected articles, 12 criteria proposed by Polit and Beck for assessing the quality of the research were used for the remaining six studies [17]. These criteria allow for an overall assessment of the objectives, sample characteristics, findings, and results of the studies. Each of these selected studies was assessed by the two researchers independently using all of the criteria. A score of '1' or '0' was assigned for each criterion according to whether or not the specified paper met the criteria. The total score that could be attributed to one paper ranged from 0 to 12. All papers with study subgroups were assessed by two researchers independently. A paper was considered eligible if a score of 7 or more was attributed after the quality assessment. Of the studies (n=9) which were assessed for quality by the independent raters, three were considered 'weak' and one was considered 'moderate'. As it was planned to include only the 'strong-quality' and 'moderate-quality' studies in the meta-analysis, 6 studies meeting those quality criteria were identified. However, one of these six studies was then excluded as it was not found suitable for inclusion in the meta-analysis due to the sample size exceeding the maximum levels and would therefore prevent the calculation of the effect size. Thus, 5 studies were included in the meta-analysis. The quality assessment of the 5 studies included in the meta-analysis and the distribution of scores attributed to each field are presented in Table 1.

Assessment of the Primary Outcomes In the papers included in the study, the assessments and analyses were performed based on the comparisons of the delirium incidence rate, age ranges, the number of risk factors and polypharmacy, gender distribution in the sample, the use of evaluation scales for diagnosing delirium, and the scores obtained from those scales between the groups developing delirium and the groups without delirium. As the studies did not have a randomized controlled design, descriptive analyses were included in the study.

Data Analysis 'Comprehensive Meta-Analysis Academic/Non-profit Pricing (Version 3)' was used as the licensed software for the data analysis. The data reported in all of the papers meeting the inclusion criteria and included in the study were entered into the CMA software, and the papers were assessed for heterogeneity. The effect sizes, study weights, 95% confidence intervals, and the overall effect size of all studies were calculated using the random effects model or the fixed effects model based on the p-values of ≤ 0.05 or >0.05 respectively, which resulted from the heterogeneity tests in the analysis of the groups. The 'RR' and 'OR' values were primarily used to evaluate the overall effect size in the analysis of the binary data. The statistical significance limit was accepted as p ≤ 0.05 for the assessment of the overall effect. A funnel plot was performed to test the publication bias, and the results obtained in the calculations of the classic fail-safe N and Tau coefficients were used.

3. Results

Descriptive Results Of the studies included in the meta-analysis, one was a descriptive study and the remaining four were cohort studies, all published between 2010 and 2019. Each of these five studies was published in one of the following years; 2012, 2014, 2016, 2017, and 2019. The oldest study was published in 2012.

Characteristics of the Study Samples in the Studies Included in the Assessment, and the Therapeutic Areas on which the Studies were Conducted. A total sample size was determined of 892 patients aged 65 years and older. The size of the sample in each study included in the meta-analysis ranged from 70 – 385. In the study by Kupeli et al (2016) included in the meta-analysis, the mean age was 71.8 ± 6.7 years in the control group and 72.7 ± 6.1 years in the patient group [5,18]. There were no statistically significant differences between the groups (p = 0.486). The percentage of female patients was 69.1% and the percentage of male patients was 30.9%. A total of 46 patients underwent hip arthroplasty, 65 patients underwent knee arthroplasty, and 9 patients underwent other types of surgery for limb amputation, treatment of fractures, and tumors. In the study by Liang et al. (2014), all patients who were scheduled to undergo orthopaedic surgery with a mean age of 74.7 ± 7.8 years were included [19]. Of those patients, 46.6% were males and 63.4% were females, 25.9% underwent scheduled spinal surgery, 43.1% knee surgery, 17.7% hip arthroplasty, and 13.4% open reduction and internal fixation. In the study by Beishuizen et al. (2017), the study population comprised 32.3% males and 77.7% females with a mean age of 86.7±6.5 years, all of whom underwent surgery for hip fracture [20]. Atay et al. (2012), reported on a sample of 54.3% females and 45.7% males with a mean age of 75.03 ± 9.82 years, hospitalized for surgical treatment of a hip fracture [21]. The study by Bert et al. (2019) included 75% females and 25% males with a mean age of 85 years, all of whom underwent surgery for femoral fractures [22].

Questionnaires and Scales used in the Studies Included in the Assessment In addition to the DSM criteria, several types of scales have been developed to screen and diagnose delirium and to determine the severity [23,24]. These scales are the confusion assessment method for the ICU (CAM-ICU), the intensive care delirium screening checklist (ICDSC), the NEECHAM confusion scale, the nursing delirium screening scale (Nu-DESC), and the delirium detection score (DDS) [25-29]. Of those evaluation tools, the most commonly studied ones are CAM-ICU and ICDSC and the most commonly used one is CAM-ICU. Of the studies included in the present meta-analysis, Bert et al. (2019) and Kupeli et al. (2016) used the confusion assessment method to assess the cognitive status of the patients [5,22]. Beishuizen et al. (2017) used the informant questionnaire on cognitive decline-short form and the mini-mental state examination (MMSE) [20,25-29]. Atay et al. (2012) used the standardized mini-mental test [21]. Liang et al. (2014) administered the Chinese geriatric depression scale, activities of daily living, confusion assessment method, MMSE, the American Society of Anaesthesiologists physical status classification (ASA), and the Lawton instrumental activities of daily living (IADL) scale [19]. For delirium screening and rating, Kupeli et al. used the Delirium Rating Scale-Revised-98 version, Atay et al. used the delirium rating scale (DRS), and Beishuizen et al. investigated S100B in blood serum samples [5,18-22].

Risk Factors Leading to Delirium and Delirium Development in the Studies Included in the Assessment The incidence of postoperative delirium after orthopaedic surgery was seen to range from 9.1-34.11%. The risk factors commonly evaluated in all of the studies were found to be age, gender, polypharmacy, and delirium assessment scale scores. The data of those risk factors were analyzed in the present study.

Methodological Quality Assessment In the present study, the inter-rater agreement between the researchers was found to be 81% based on the quality assessment score. The reliability analysis revealed that Cohen's kappa was 0.84, which was within the 95% confidence interval [Confidence Interval (CI): 0.767-0.873]. Kappa values of <0 indicate poorer agreement than expected by chance, values in the range between 0.01 - 0.20 indicate negligible agreement, 0.21 - 0.40 indicate slight agreement, 0.41 - 0.60 indicate moderate agreement, 0.61 - 0.80 indicate substantial agreement, and values in the range of 0.81 - 1.00 indicate perfect agreement [8], or values of ?0.75 indicate excellent, 0.40-0.75 indicate moderate-good, and values less than 0.40 are considered poor agreement [9]. The kappa value of 0.82 in the present study indicates perfect inter-rater agreement. The mean scores attributed to each study by the investigators are presented in Table 1.

Results of the Analysis In geriatric patients; who had undergone orthopaedic surgery and developed delirium, the following high-level, in the positive direction, and statistically significant overall effect size of 5.21 (CI; 1.33-20.33; p <0.05) was found for gender, 1.33 (CI; 0.58-2.06; p 0.05) was found for age, and 11.30 (CI; 4.70-27.12; p 0.05) was found for polypharmacy, and the following low-level, in the positive direction, and statistically significant overall effect size of 0.12 (CI; 0.05 -0.27; p 0.05) was found for MMSE as the risk factors leading to the development of delirium. A forest plot chart was used to present the results of the meta-analysis of gender, age, polypharmacy, and the delirium screening scores in the five included studies. o Gender The heterogeneity test for the gender variable as a risk factor of delirium yielded a p-value of < 0.05 and a Q-value (29.509) of more than the corresponding df value revealing a heterogeneous construct (Table 2). The statistical I2 value was calculated as 86.44. As a result of the calculations, the study effect sizes for gender were evaluated according to the random effects model.

The random effects analysis showed that the overall effect size of gender on the development of delirium was at a high-level, positive, and statistically significant with a value of 5.21 (CI: 1.33-20.33; p<0.05). Forest plot of analysis for gender is shown in Figure 2. Female gender was found to play a more significant role compared to male gender as a major risk factor for the development of delirium. o Age The heterogeneity test yielded a p-value of < 0.05 and Q-value (51.866) more than the corresponding df value (Table 2). The examination of the included studies in the meta-analysis for the age variable as a risk factor for delirium showed that the studies had a heterogeneous construct.

The statistical value I2 was calculated as 92.28. As a result of the calculations, the distribution of the effect sizes was evaluated using the random effects model. Female gender was found to play a more significant role compared to male gender as a major risk factor for the development of delirium.

According to the analysis carried out according to the random effects model, the overall effect size of age on delirium development was found to be 1.33 (CI: 0.58-2.06; p <0.05) as a high value with a positive direction and statistical significance. Forest plot of analysis for age is shown in Figure 3. This observed effect was higher especially in patients aged ≥75 years compared to patients ≤74 years and the risk for delirium was seen to increase with increasing age.

o Polypharmacy The heterogeneity test yielded a p-value of >0.05 and Q-value (0.230) less than the corresponding df value (Table 2). The meta-analysis of the included studies showed that the studies examined for the polypharmacy variable as a risk factor for delirium had a homogeneous Effect size and 95% interyal Test of null (2-Tail) Heterogenity Tau-squared Variable Number

Studies

Point estimate

Lower

limit

Upper

limit

Z-value P-value Q-value df (Q) P-value I-squared Tau Squared

Standard

Error

Variance Tau Gender 5 5.425 3.283 8.964 6.600 0.000 29.357 4 0.000 86.374 2.083 1.708 2.916 .,443 Age 5 0.899 0.729 1.070 10.343 0.000 51.866 4 0.000 92.288 0.630 0.562 0.316 0.794 Polypharmacy 3 11.300 4.708 27.127 5.427 0.000 0.230 2 0.891 0.000 0.000 0.762 0.581 0.000 Delirium screening score

3 0.125 0.081 0.192 -9.419 0.000 13.987 4 0.007 71.402 0.000 0.762 0.581 0.000 construct. The statistical I2 value was calculated to be 0.000. As a result of the calculations, the effect size distributions were evaluated using the fixed effects model.

According to the analysis carried out according to the fixed effects model, the overall effect size of polypharmacy on developing delirium was 11.30 (CI: 4.70-27.12; p <0.05) and it was found to be at a high level with a positive direction and statistical significance. Forest plot of analysis for polypharmacy is shown in Figure 4. This effect was determined to be higher in patients using three or more types of drugs compared to patients using fewer than three drugs concurrently and the risk of delirium increased as the number of medications increased.

o Delirium Screening Score The heterogeneity test yielded a p-value of < 0.05 and a Q-value (13.937) more than the corresponding df value. The meta-analysis of the included studies for the delirium screening score variable as a risk factor for delirium showed that the examined studies had a heterogeneous construct. The statistical value of I2 was 0.611. As a result of the calculations, the effect size distributions were evaluated according to the random effects model. According to the results of the analysis carried out using the random effects model for the heterogeneity test to investigate the delirium screening score variable, the overall effect size of the MMSE scores on developing delirium was found to be 0.12 (CI: 0.05-0.27; p <0.05) at a low level and with a positive direction and statistical significance. Forest plot of analysis for delirium screening score is shown in Figure 5. Compared to patients with MMSE scores of 24-30, especially in patients with MMSE scores of 17-23, the observed effect was higher, and the risk of delirium increased as the MMSE scores decreased.

In the studies examined within the scope of this research, the type of surgery, whether hip or knee surgery, was not determined to play a moderator role (Qb = 3.58; p> 0.05). Performing the surgical intervention either in the hip (r = 0.39) or the knee (r = 0.36) changed the effect size of the relationship between developing delirium and the risk factors of gender, age, polypharmacy, and delirium screening scores.

Publication Bias The funnel scatter plot is considered a demonstration of the possibility of the publication bias. In the funnel scatter plot, the standard error values of the included studies are shown on the y-axis and the effect size is shown on the x-axis. Studies with small standard error values are scattered towards the top of the funnel shape and close to the average effect size. The studies with a large standard error value scatter towards the bottom of the Figure 5. In the absence of any publication bias, the studies are expected to scatter symmetrically on both sides of the vertical line showing the combined effect size [29].

We could not observe publication bias according to funnel plots for the variables that we analyzed. Another way to determine the publication bias in meta-analysis research is to calculate Kendall's tau-b coefficient. In the absence of any publication bias, this coefficient is expected to be close to 1 and the double-tailed p-value is expected to be greater than 0.05 not indicating a significant difference [30]. According to the results of this statistical analysis (Kendall's tau-b=1.09; p=0.512), publication bias was not detected in the studies included in this meta-analysis.

4. Discussion

Delirium is a cognitive disorder commonly seen in elderly patients who undergo orthopaedic surgery, and results in a longer hospital stay with direct effects on morbidity and mortality rates. It has been previously reported that preoperative presence of cognitive impairment is one of the major risk factors for postoperative delirium after hip fracture surgery [31].

Another study, which aimed to predict the development of delirium during the preoperative care of orthopaedic surgery patients, found that patients with MMSE scores of ≤23 had a higher risk of developing postoperative delirium [32]. In this meta-analysis, patients with MMSE scores of 17-23 were seen to be at greater risk than those who scored 24-30.

Previous studies investigating postoperative delirium in adult patients have reported that polypharmacy could precipitate delirium and is an independent risk factor [33, 34]. In this meta-analysis, polypharmacy, especially when three or more drugs were used, was found to be a risk factor for the development of delirium.

In a study of elderly patients scheduled to undergo major orthopaedic surgery, male gender was reported to be a risk factor for delirium [35,36]. However, another study could not determine any significant effects of the gender variable on delirium [37]. In this meta-analysis, the effect of gender on precipitating delirium was significant, with female gender in particular showing a greater effect as a risk factor.

A study of adult intensive care patients reported old age as a risk factor for developing delirium [34]. Other studies on elderly patients in the postoperative period also reported that advanced age was a risk factor for delirium [38,39]. In this meta-analysis, age >75 years was observed to be a risk factor for the development of delirium after surgery.

It has been reported that the administration of a multi-factor intervention program based on care and supportive treatment in geriatric hip fracture patients starting from the time of admission reduced the incidence of delirium at admission from 34% to 22% [32].

It may not be possible to completely prevent the development of delirium in patients. However, a comprehensive evaluation of the patient can allow for the identification of high-risk patients and any further deterioration of the patient may be prevented [40,41].

The level of evidence of the development of postoperative delirium in elderly patients after orthopaedic surgery is quite limited throughout the world. Almost all of the studies have attempted methods to prevent delirium, but no studies could be found which have reported only the risk factors. Despite several available tools for the assessment of cognitive functions, no comprehensive tools have been developed to include all types of risk assessments in order to predict the emergence of delirium. Furthermore, studies have continued to stress the inadequacies of the incomplete information and practice reported in the literature for the use by health personnel working in the field and for the early recognition of delirium [5,15]. In addition to the MMSE examined in this study, several major dementia screening and rating scales are in current use, including IADL, Blessed Dementia Rating Scale- Consortium to Establish a Registry for Alzheimer's Disease (BDRS-CERAD), physical self-maintenance scale (PSMS), and the Katz index of independence in activities of daily living.

As dementia is a medical syndrome, collaboration with the psychiatry team and clinical evaluation are critical for the diagnosis. It will be more beneficial to remember the importance of collaboration and the importance of the clinical evaluation in addition to the use of the dementia screening and rating scales. The significance of the clinical evaluation and collaboration with psychiatry should be recognized in comorbid conditions of dementia and delirium and in conditions, where making a differential diagnosis is difficult.

5. Conclusions

Considering that delirium is an important cognitive problem throughout the world, it can be considered that this meta-analysis, which was conducted to identify the delirium risk factors and contribute to the provision of care and treatment in delirium, will pave the way for future studies. Within this scope, there can be seen to be a clear need for experimental studies, especially those providing adequate levels of evidence about this particular subject matter.

Orthopaedic surgery patients should be evaluated for delirium in the early postoperative period, associated risk factors should be identified, and prevention strategies should be developed. Based on the common risk factors identified in the present meta-analysis, an assessment scale can be developed to predict the development of delirium in the postoperative period.