Study reference

Study characteristics

Patient characteristics ²

Intervention (I)

Comparison/ control (C) ³

Follow-up

Outcome measures and effect size ⁴

Comments

Bub, 2022

Type of study:

Retrospective chart review study

Setting and country:

Large urban, academic tertiary centre, located in the greater New York metropolitan area

Funding and conflicts of interest: Reported

Inclusion criteria:

Patients were enrolled when they were 65 years of age and older and admitted to the hospital with a low-energy (fragility) hip-fracture

Exclusion criteria:

Patients younger than 65 years old, pathologic hip fractures, periprosthetic fractures, and chronic opioid users were excluded.

N total at baseline:

Intervention: 162

Control: 128

Important prognostic factors²:

For example

age ± SD:

I: 84.11 ± 8.4

C: 83.5 ± 9.0

Sex:

I: 34.6% M

C: 22.7% M

Groups comparable at baseline?

No, the two groups were comparable for CCI, surgery type, and anaesthesia, although there were significantly more intertrochanteric fractures and fewer women in the intervention group.

Describe intervention (treatment/procedure/test):

The intervention consisted of the AGS CoCore: Ortho program. A dedicated hospitalist (eg trained geriatrics comanager), with geriatrics oversight, comanaged all patients throughout the perioperative period. Team members from orthopaedics, hospital medicine, nursing, and geriatrics rounded together 2 times a day, in the morning and again on nightly rounds. The teams discussed medical and surgical management as well as psychosocial and discharge issues.

Additionally, evidence based standardized pre- and postoperative protocols were adapted and implemented in the HER, preoperative risk stratification, and postoperative delirium prevention and discharge planning was implemented in the intervention group.

Describe control (treatment/procedure/test):

Before the implementation of comanagement, patients with hip fracture were admitted to either orthopaedics or hospitalist service with the other on board as a consultant. Patients with an American Society of Anesthesiologists classification I and II were admitted to orthopaedics, whereas those over III were admitted to the hospitalist service. American Society of Anesthesiologists class III classification was admitted to either orthopaedics or hospitalists depending on the surgeon’s comfort level. There was no formal agreement between orthopaedics and hospital medicine and no standard process for preoperative assessment or perioperative care (eg, pain management).

Length of follow-up:

No follow up.

Loss-to-follow-up:

NA

Incomplete outcome data:

NA

Outcome measures and effect size (include 95%CI and p-value if available):

Surgical metrics

Time to OR significant lower in I group (I 30.3 hours ± 23.4 vs C 36.2 ± 23.7, p=0.026*).

No difference in operative time (I 60.6 ± 26.2 vs C 61.5 ± 23.6, p=0.716).

Lower length of stay (I 5.0 days ± 2.8 vs C 6.8 ± 6.8, p=0.005*).

Opioid use lower in I group (I 11.1 MME/d ± 14.6 vs C 16.0 ± 17.3, p=0.011*).

Perioperative testing

No difference in preoperative transthoracic echocardiogram (I 21 (13.0%) vs C 28 (21.9%), p=0.58).

Discharge disposition No difference in discharge destination (p=0.197, in-hospital mortality I 2 (1.2%) vs C 4 (3.1%); skilled nursing facility I 64 (39.5%) vs 36 (28.1%); subacute rehabilitation I 79 (48.8%) vs C 67 (52.3%); home I 14 (8.6%) vs C 17 (13.3%); hospice/specialized care I 3 (1.9%) vs C 4 (3.1%).

Perioperative complications

No difference in preoperative and postoperative transfusion.

No difference in failed trial of voiding and discharged with Foley.

I group had decreased use of indwelling bladder catheters preoperative (I 46.9% vs C 68.0%, p<0.001*) and postoperative (I 58.0% vs C 83.6, p<0.001*).

I group had decreased incidence of urinary tract infections (I 21 (13.0%) vs C 36 (28.1%), p=0.002*).

No difference in deep vein thrombosis (I 1 (0.6%) vs C 1 (0.8%), p=1.000).

No difference in aspiration pneumonia (I 15 (9.3%) vs C 12 (9.4%), p=0.703).

No difference in pulmonary embolism (I 1 (0.6%) vs C 0, p=1.000).

No difference in cerebrovascular accident (I 0 vs C 2 (1.6%), p=0.194).

No difference in myocardial infarction (I 1 (0.6%) vs C 0, p=1.000).

No difference in return to operating room (I 1 (0.6%) vs 0, p=1.000).

No difference in 6-month mortality rate (I 10 (6.2%) vs C 8 (6.3%), p=1.000).

Less complications in the I group (I 34 (21.0%) vs C 44 (34.4%), p=0.012*).

No comments

Fan, 2021

Type of study:

Pre-post retrospective study

Setting and country: single trauma centre in China

Funding and conflicts of interest: Reported

Inclusion criteria:

(1) acute intertrochanteric fractures (time from injury to admission was limited to 3 weeks); (2) age ≥60 years old; and (3) low-energy injury. A low-energy injury was defined as an injury which patients would sustain while falling over slippery ground in a walking or sitting position.

Exclusion criteria:

Pathological fracture; high-energy injury, such as car crash; peri-prosthetic fracture; multiply traumatized patients or terminal malignancies.

N total at baseline:

Intervention: 241

Control: 249

Important prognostic factors²:

For example

age ± SD:

I: 79.9 ± 8.1

C: 78.8 ± 7.2

Sex:

I: 30.52% M

C: 31.95% M

Groups comparable at baseline?

Yes

Describe intervention (treatment/procedure/test):

The MDT model involved orthopaedic surgeons, geriatricians, anaesthesiologists, specialists of the intensive care unit (ICU), and physiotherapists. This model consisted of a pathway of care spanning Emergency Department (ED) presentation to discharge from hospital. The traumatologist
decided the suitability of the surgical treatment, technique to use, and when progressive weight-bearing could begin. The geriatrician managed comorbidities and polypharmacy to make patients clinically stable and ready for surgery, to reduce peri-operative complications and promote early functional recovery. A preoperative anaesthesia evaluation was performed for risk
assessment by an anaesthesiologist within 12 h in the ED.
Postoperative observation for 24–48 h in ICU was planned for
high-risk patients. Furthermore, early geriatric rehabilitation was
carried out by a physiotherapist. This team aimed to perform
surgical treatment within 48 h upon admission and to achieve
early discharge.

Additionally to the intervention, suitability of surgical treatment, preoperative anaesthesia evaluation, post-operative evaluation in ICU, early geriatric rehabilitation, and early surgery and early discharge were also implemented in the intervention group.

Describe control (treatment/procedure/test):

Traditional Orthopaedic Care model (TOC): When a patient suspected of having a fragility hip fracture was seen in the ED, an initial X-ray of the injured hip was obtained. Once the intertrochanteric fracture diagnosis was confirmed, the patient was admitted to the orthopaedic ward under the care of an orthopaedic surgical team. Treatment was generally managed by trauma surgeons and their team, who had no specific geriatric expertise. The geriatric assessment was not performed during inpatient treatment. Specialist consultants were called on according to their perception of patients’ clinical conditions. The preoperative anaesthesia assessment was performed on the day before surgery. Surgical treatment was performed by trauma surgeons. The patient would be transferred to the department of rehabilitation medicine after discharge.

Length of follow-up:

One months after discharge

Loss-to-follow-up:

Total: 23 patients were excluded, but reasons were not reported.

Outcome measures and effect size (include 95%CI and p-value if available):

Time to surgery

Significantly lower in I group: I 1.7 ± 1.3 days vs C 2.4 ± 1.5 days, p< 0.001*.

The proportion of patients receiving surgery within 24 h (I 148 (61.4%) vs. C 87 (34.9%), p <0.001*) and 48 h (I 195 (80.9%) vs. C 158 (63.5%), p < 0.001*) after admission to the ward was significantly higher in the I group compared with those in the C group.

Total length of stay

Lower in I group than C group (I 4.0 ± 2.5 days vs C 5.0 ± 2.8 days, p< 0.001*).

Anaemia

No difference in proportion of patients receiving red blood cell transfusion (I 80 (33.2%) vs C 75 (30.1%), p=0.464).

The proportion of patients with a haemoglobin >11 g/dl was higher in the I group (I 58 (24.1%) vs C 37 (14.9%), p=0.010*).

Postoperative complications

Patients in the I group had significantly lower proportion of postoperative complications (I 61 (25.3%) vs. C 110 (44.2%), p < 0.001*)

Lower rates of deep venous thrombosis (I 18 (7.9%) vs. C 32 (12.9%), p = 0.049*)

Lower rates of pneumonia (I 9 (3.8%) vs. C 20 (8.0%), p = 0.045*)

Lower rates of delirium (I 10 (4.1%) vs. C 21 (9.2%), p = 0.025*).

No statistically significant differences were observed in wound infection (I 3 (1.2%) vs C 4 (1.6%), p=0.736), urinary tract infection (I 6 (3.5%) vs C 10 (4.0%), p=0.342), cerebral vascular accident (I 5 (2.1%) vs C 6 (2.4%), p=0.802), acute coronary syndrome (I 8 (3.3%) vs C 13 (5.2%), p=0.299), and gastrointestinal bleeding (I 2 (0.8%) vs C 4 (1.6%), p=0.435).

30-day Harris Score did not differ (I 80.8 ± 7.5 vs C 80.3 ± 7.0, p=0.440).

Mortality

No difference in in-hospital mortality (I 1 (0.4%) vs C 2 (0.8%), p=0.582).

No difference in 30-day mortality (I 4 (1.7%) vs C 6 (2.4%), p=0.557).

No comments

Giannotti, 2022

Type of study: pre-post retrospective study

Setting and country: single-centre study based at the oncological surgery of Policlinico San Martino of Genoa, Italy

Fundings and conflict of interest: not reported

Inclusion criteria: Patients aged 70 or older admitted for elective GI cancer surgery or palliative treatments and required a hospital stay of at least 1 day.

Exclusion criteria: if patients had any instability needing acute surgery, or if they were admitted for secondary surgeries or because of a postoperative complication.

N total at baseline:

Intervention: 90

Control: 117

Important prognostic factors

age median (IQR):

I: 79.0 (76.0-83.0)

C: 82.0 (78.0-85.0)

Sex:

I: 53.3% M

C: 64.1%

Groups comparable at baseline?

Patients in control group are slightly older. Comparable in terms of cancer diagnosis and metastatic disease.

Describe the Intervention:

Model of geriatric care consisted of initiation of daily targeted geriatrician-led ward rounds focusing on older patients with cancer. During the inpatient postoperative period, patients were followed by the same geriatrician in a consulting role, with the surgical team in a primary role. The geriatric comanagement group included a daily board round led by a geriatrician who discussed the care management during the clinical sessions.

Patients received a pre-operative CGA and a frailty assessment.

Perioperative phase followed the major principles of the Enhanced Recovery After Surgery (ERAS)

Describe the control:

Older patients with cancer underwent a preoperative comprehensive geriatric assessment (CGA) and frailty assessment to stratify patients’ frailty and performance status within 2 weeks before admission to the surgical department by a geriatrician working at the geriatric clinic of our hospital. This CGA was aimed at identifying high risk patients and the assessment was followed by recommendations based on the identified health issues. Referral to the geriatrics service was based on the surgical team preference and clinical judgment, but not based on a formal frailty screening tool. During hospitalization and perioperative phase, patients from the control group were assessed daily by the surgical team and medical consultants were called in as needed.

Perioperative phase followed the major principles of the Enhanced Recovery After Surgery (ERAS).

Length of follow up

One year

Loss to follow up

No missing data reported

Outcome measures and effect size (include 95%CI and p-value if available):

Post-operative complications

Fewer rates of bacteriemia/sepsis in the I group (I 6.7% vs. C 20.5%; p 0.011)

Fewer rates of haematological complications in the I group (I 7.8% vs. 19.7%; p 0.031). Similar rates of acute kidney failure, pneumonia, cardiovascular complications and neurological complications.

Patients in the geriatric comanagement group lower risk of grade I-V postoperative complications OR 0.43; 95% CI 0.25-0.73; p < 0.002)

Delirium

Lower 4AT scores in the I group (mean (SD)): I 1.44 (2.62) C 2.19 (3.27); p 0.03

Rehospitalization

Lower rates of rehospitalization within 1 year in the I group (I 23.5% vs C 35.9%; p 0.006).HR 0.49; 95% CI 0.30-0.81; p < 0.006)

Length of stay

No difference in median length of stay or overall survival between groups.

Hafner, 2021

Type of study:

Retrospective cohort study

Setting and country: Single setting study based on data collected from the electronic patient records at the university hospital of Aachen, Germany.

Funding and conflicts of interest:

Reported

Inclusion criteria:

All geriatric patients aged > 70 years who were admitted with lumbar spine, pelvic, or acetabular fractures in the period from January 2012 to September 2019 were included.

Exclusion criteria:

Polytrauma patients (Injury Severity Score ≥ 16),

tumour-associated fractures, and patients treated only in the ICU were excluded.

N total at baseline:

Intervention: 224

Control: 137

Important prognostic factors²:

For example

age ± SD:

I: 82.2 ± 6.2

C: 81.5 ± 6.2

Sex:

I: 30.4% M

C: 25.5% M

Groups comparable at baseline?

No difference in age and sex. Differences in ASA classification a larger number of patients in the OGC cohort were assigned to ASA group III/IV (UC: 68.8% vs. OGC: 75.4%; p = 0.002). Moreover, the two cohorts differed significantly regarding the fracture entities. While pelvic fractures were the most common fracture
entity in the OGC cohort (43.8%), they ranked second in the UC cohort with 37.2%, after lumbar fractures, with 40.1%. However, a post hoc test revealed that statistical significance was limited to the subgroup of others only (UC: 6.6% vs. OGC: 0.4%; p < 0.001).

Describe intervention (treatment/procedure/test):

The ortho-geriatric model implemented in this study is based on the ward round model in conformity with the certification guidelines of the DGU® [12,19]. Patients were admitted to and treated at the trauma surgery ward with the routine consultation of a geriatrician in an interdisciplinary ward round twice a week and an interdisciplinary team conference once a week. Moreover, representatives of nursing, occupational therapy, physiotherapy, and case management took part in the treatment process from admission until discharge. All participants were advised to provide care following the standard operating procedures (SOPs) provided by the DGU® for geriatric trauma patients.

Geriatric assessment was also performed.

Describe control (treatment/procedure/test):

Usual care, before implementation of the ortho-geriatric model.

Length of follow-up:

No follow-up.

Outcome measures and effect size (include 95%CI and p-value if available):

Clinical complications

Higher rate of urological complications in the I group (I 84 (37.5%) vs C 35 (25.5%), p=0.021*).

No difference in total complications (I 150 (67.0%) vs C 80 (58.4%), p=0.100).

Significant fewer numbers of revision surgery in the I group (I 7 (3.1%) vs C 3 (2.2%), p=0.012*).

No difference in dehydration, electrolyte disorder, anaemia, transfusion, pulmonary complications, delirium, cardiac complications, gastrointestinal complications, surgical complications, neurological complications, acute kidney injury, sepsis/SIRS/shock, complications during anaesthesia, in-hospital fall, and in-hospital mortality (all p>0.05).

Secondary outcome measures

No difference in total length hospital stay (I 10.4 ± 8.3 vs C 10.4 ± 5.3 days, p=0.111).

Increase in time to surgery in I group (I 112.2 ± 75.8 vs C 72.4 ± 54.5 hours, p<0.001*).

Higher amount of patients with suspected dementia (I 13 (5.8%) vs C1 (0.7%), p=0.022*).

Increased osteoporosis therapy at discharge (p<0.001*).

More patients were mobilized on the first day in the I group (I 101 (86.3% in 117 patients) vs 48 (57.1% in 84 patients), p<0.001*).

No difference in change in walking ability before and after the fracture (p=0.252).

No difference in therapy regimen (surgery or conservative) (p=0.255).

No difference in time at ICU (I 11.2 ± 36.0 vs C 14.3 ± 47.4 hours, p=0.660).

No difference in osteoporosis therapy at discharge (I 19 (9.4% in 203 patients) vs C 16 (11.7% in 137 patients).

No comments.

Kalmet, 2019

Type of study:

Retrospective cohort study

Setting and country: University hospital in Maastricht, the Netherlands

Funding and conflicts of interest:

Reported

Inclusion criteria:

Patients aged 65 years or older with a surgically treated low-energy fracture

Exclusion criteria:

Patients with a high-energy hip fracture, patients with several or complicated fractures, and patients not living in the hospital area were excluded.

N total at baseline:

Intervention: 182

Control: 216

Important prognostic factors²:

For example

age ± SD:

I: 83.4 ± 7.4

C: 82.2 ± 7.5

Sex:

I: 29.1% M

C: 29.2% M

Groups comparable at baseline?

Patients in the C group had more comorbidities and were more likely to have a femoral neck fracture. No differences in gender, age, or ASA classification

Describe intervention (treatment/procedure/test):

The intervention consisted of a multidisciplinary team (MCP). This MCP consists of an orthopaedic trauma surgeon, a geriatrician, an anaesthesiologist, and a physiotherapist. These
disciplines are all actively involved in the decision making process regarding the care that patients need from the first
presentation at the emergency department until they are discharged from the hospital. Additional medical specialties
remain available for consultation depending on the comorbidities of the patient. The aim of the team is to perform surgical treatment within 24 hours upon admission and to achieve discharge within 4 days. To achieve this goal, agreements have been set in place with rehabilitation facilities to transfer the surgically treated patients to a patient-centred destination as soon as possible. This may either be a rehabilitation centre or nursing home with rehabilitation facilities. The postoperative protocol for both groups, MCP and UC, was early mobilization and early full weight bearing.

Additionally, patients in the intervention group received early surgery and early discharge.

Describe control (treatment/procedure/test):

The comparison consisted of usual care. Usual care protocol includes standard traditional treatment by an orthopaedic trauma surgeon at the trauma unit with a follow-up at the out-patient clinic. Physiotherapy is prescribed when the patient is discharged home.

Length of follow-up:

At least two years

Loss-to-follow-up:

49.9% of patients were still alive after at least 2 years follow-up. 14 of the remaining 159 patients could not fulfil the questionnaire due to cognitive impairment. The response rate of the questionnaire was 65.6% (95 of 145 participants).

Outcome measures and effect size (include 95%CI and p-value if available):

In-hospital outcomes

Mean time to surgery was shorter in the I group (I 18.2 ± 9.3 hours vs C 25.3 ± 13.9 hours, p<0.01*).

Number of patients who had to wait more than 24 hours for surgery was lower in the I group (I 17.6% vs C 44.9%, p<0.01*).

Longer mean length of hospitalization in I group (I 15.1 ± 15.7 vs C 12.3 ± 7.3, p=0.02*).

Patient reported outcomes

No difference in mean QoL (I 47.9 ± 24.4 vs C 45.4 ± 27.6, p=0.65), physical QoL (I 36.3 ± 28.6 vs C 35.8 ± 28.9, p=0.93), or mental QoL (I 59.5 ± 25.0 vs CG 54.9 ± 31.0, p=0.45).

No difference in pain 30-day after hip fracture (I 4.2 ± 2.8 vs C 3.9 ± 2.7, p=0.56), 1-year after fracture (I 2.7 ± 2.8 vs 1.9 ± 2.7, p=0.17), and at follow-up questionnaire (I 2.8 ± 4.0 vs 1.8 ± 2.7, p=0.19).

Complications and mortality outcome measures

Lower incidence of postoperative complications in I group (I 82 (45.1%) vs C 179 (82.9%), p<0.01*).

Lower incidence of postoperative delirium (I 35 (19.2%) vs C 98 (45.4%), p<0.01*).

No difference in 30-day mortality (I 15 (8.2%) vs C 17 (7.9%), p=0.90) and 1-year mortality (I 61 (33.5%) vs C 80 (37.0%), p=0.50).

No comments.

McDonald, 2018

Type of study:

Prospective cohort study

Setting and country:

Duke university hospital USA

Funding and conflicts of interest:

Reported

Inclusion criteria:

All surgical candidates 85 years and older undergoing elective colorectal, general, and hepatopancreaticobiliary surgical procedures (these were at higher risk). Patients between age 65 and 84 years were considered eligible when any 1 of the following conditions was present: prior diagnosis of a cognitive disorder, weight loss of more than 4.54 kg in the last year, multimorbidity (presence of 2 or more chronic medical conditions), polypharmacy (more than 5 prescription medications), any visual or hearing impairment, or if the surgeon perceived an increased estimation of risk.

Exclusion criteria:

This analysis excluded participants who underwent outpatient surgical procedures.

N total at baseline:

Intervention: 183

Control: 143

Important prognostic factors²:

For example

age ± SD:

I: 75.6 ± 6.8

C: 71.9 ± 6.4

Sex:

I: 46.6% M

C: 51.0% M

Groups comparable at baseline? Patients in the I group were older, more often white, more often an active smoker, had higher haemoglobin levels, and more comorbid conditions

Describe intervention (treatment/procedure/test):

The POSH preoperative assessment team included a geriatrician, geriatric resource nurse, social worker, program administrator, and nurse practitioner
from the Preoperative Anesthesia Testing clinic to complete a comprehensive preoperative geriatric evaluation. The team actively engaged patients and their families in preoperative risk assessment and modification, focusing on specific care points considered crucial for optimizing care: cognition, medications, comorbidities, mobility, functional status, nutrition, hydration, pain, and advanced care planning. During the preoperative visit with all health care professionals, which typically lasted 60 to 120 minutes, the POSH team offered recommendations for risk-reducing strategies in the preoperative and postoperative periods as well as anticipating needs at discharge. Discussions also included consideration of nonsurgical management options, which for some patients aligned more closely with their personal goals. In the postoperative period, the hospital geriatrics consult team followed patients daily. Physicians who conducted preoperative POSH evaluations also participated in rounds on the inpatient geriatrics consult service. To facilitate implementation of recommendations made before surgery, the inpatient geriatrics team collaborated with the surgical teams, assisting with the management of medications, chronic medical conditions, pain, and recognition and treatment of common postoperative complications, including delirium. The geriatrics and surgery teams also jointly counselled patients and families, helping them prepare for discharge and posthospital care.

Additionally, patients in the intervention group received preoperative risk assessment and modification.

Describe control (treatment/procedure/test):

Control cohort of adults older than 65 years who underwent similar surgical procedures as determined by Current Procedural Terminology code, performed by the same surgical group, between January 2010 and May 2011.

Length of follow-up:

30-days

Loss-to-follow-up:

Intervention:

N (%): 3

Reasons (describe): NR

Control:

N (%): 1

Reasons (describe): NR

Incomplete outcome data:

No incomplete outcome data

Outcome measures and effect size (include 95%CI and p-value if available):

Health care service use

Shorter median LOS in I group (I median 4.0 d range 1-75 vs C median 6.0 d range 1-60, p<0.001*, 95%CI 1.1;4.2).

Lower 7-day all-cause readmission rates (I 5 of 180 (2.8%) vs C 14 of 142 (9.9%), p<0.001*, 95%CI 0.09;0.74).

Lower 30-day all-cause readmission rates (I 14 of 180 (7.8%) vs C 26 of 142 (18.3%), p>0.001*, 95%CI 0.19; 0.75).

Patients in the I group returned home with self-care more frequently (instead of discharge with skilled services) (I 114 of 183 (62.3%) vs C 73 of 143 (51.1%), p=0.04*, 95%CI 1.02; 2.47).

No difference in usage of health service when returning to home (I 32 of 183 [17.5%] vs C 34 of 143 [23.8%]; P = 0.16; 95% CI, 0.39-1.17).

No difference in discharge to facility (I 26 of 183 [14.2%] vs C 27 of 143 [18.9%]; P = 0.26; 95% CI, 0.39-1.28).

Postoperative complications

Lower rate of complications in the I group (I 82 of 183 (44.8%) vs C 84 of 143 (58.7%),

p=0.01*, 95%CI 0.37; 0.89).

Lower incidence of postoperative cardiogenic or hypovolemic shock in I group (I 4 of 183 [2.2%] vs C 12 of 143 [8.4%];P = <.001*; 95% CI, 0.08-0.77), bleeding during and after surgery (I 11 of 183 [6.1%] vs C 22 of 143 [15.4%]; P = <.001*; 95% CI, 0.16-0.75), and postoperative ileus (I 9 of 183 [4.9%] vs C 29 of 143 [20.3%]; P < .001*; 95% CI, 0.09-0.45).

Higher rates in I group of nausea/vomiting (I 25 of 183 [13.7%] vs C 5 of 143 [3.5%];P = <.001*; 95% CI, 1.62-11.71) and higher rates of documented delirium (I 52 of 183 [28.4%] vs C 8 of 143 [5.6%]; P = <.001*; 95% CI, 3.06-14.65).

No comments

Natesan, 2022

Type of study:

Retrospective cohort study

Setting and country: tertiary hospital in Singapore

Funding and conflicts of interest:

Not reported

Inclusion criteria:

Eligible patients included those who were acutely admitted to vascular surgery, aged 65 years, had 1 or more geriatric syndromes (dementia, a previous history of delirium, falls, postural hypotension) and 1 or more comorbidities (diabetes mellitus, ischemic heart disease, old stroke disease, chronic kidney disease), and were scheduled for vascular surgery or endovascular interventions. The types of vascular surgical procedure included major amputation, minor amputation, lower limb angioplasty, lower limb angioplasty with minor amputation, peripheral arterial bypass, and abdominal aortic surgery (open and closed).

Exclusion criteria:

NR

N total at baseline:

Intervention: 198

Control: 135

Important prognostic factors²:

For example

age (range):

I: 77 (63-94)

C: 76 (65-93)

Sex:

I: 53% M

C: 54% M

Groups comparable at baseline? Yes for age, gender, and operative procedure. However, they were not comparable for preoperative comorbidities, operative procedures.

Describe intervention (treatment/procedure/test):

Patients who fulfilled the criteria were referred to a geriatric surgical service. The geriatric team assessed the patients by performing a comprehensive geriatric assessment, identifying major contributors of frailty (cognitive impairment, low mood, malnutrition), optimizing the risk factors for delirium, and helping the patient and family with shared decision making for surgical intervention and postoperative discharge planning. In the preoperative period, each patients was seen 2-3 days before surgery by the geriatric team, to optimize delirium, depression and anxiety, and malnutrition. In the postoperative period, the geriatric team observed the patients postoperatively on a daily basis (5 days a week, Monday-Friday) to optimize comorbidity treatments and address postoperative complications early. The geriatric team helped the surgical team manage medications on a daily basis. The geriatric team also managed postoperative pain relief and discontinued unnecessary intravenous lines and urinary catheters. The team also encouraged the patient to sit out of bed and made early referrals to physiotherapy. An advanced practice geriatric nurse on the geriatric surgical team spoke with family members to identify social support and caregiver stress for discharge planning. She also identified the rehabilitation potential of each patient through consultations with therapists and coordinated early transfers to community hospitals. Suitable cases
were referred to day care or dementia day care centres, and socially complex and dependent cases were referred to social workers to arrange nursing home placement.

Additionally, patients in the IG received a comprehensive geriatric assessment, delirium prevention, shared decision making support, depression and anxiety prevention, and malnutrition prevention

Describe control (treatment/procedure/test):

Retrospective cohort of 135 vascular patients without geriatric surgical service comanagement from January 2014 to July 2015 who would have fulfilled the criteria for geriatric consultation.

Length of follow-up:

30-days

Loss-to-follow-up:

NR

Incomplete outcome data:

NR

Outcome measures and effect size (include 95%CI and p-value if available):

Length of stay

Lower length of stay in the I group (I 11.6 days vs C 20.8 days, p=0.001*).

Complications

No difference in postoperative delirium (I 22 (11%) vs C 14 (10%), p=n/s).

No difference in fluid overload (I 6 (3%) vs C 10 (7%), p=0.073).

No difference in acute myocardial infarction ((I 11 (6%) vs C 8 (6%), p=n/s).

No difference in nosocomial pneumonia (I 16 (8%) vs C 19 (14%), p=0.101).

Less nosocomial urinary infections (I 6 (3%) vs C 16 (12%), p=0.003).

Readmission

Lower 30-day readmission (I 44 (22%) vs C 46 (34%), p=0.011).

Article also describes results for an ‘early’ period, directly after implementation of the intervention, but these results are not reported here. This because other studies also have a washout period.

Pernik, 2021

Type of study:

Retrospective cohort study

Setting and country:

USA

Funding and conflicts of interest:

Reported

Inclusion criteria:

Patients with an increased perioperative risk of adverse outcomes were included.

Exclusion criteria:

Patients who underwent non-spinal procedures, emergency, outpatient, or observational (<24 hour stay) surgery were excluded. We also excluded procedures involving only lumbar laminectomy without fusion, anterior cervical discectomy without myelopathy, spinal cord stimulator placement or replacement, arteriovenous fistula resection, and spinal cord or column tumours, as these cases are not referred to UTSW POSH.

N total at baseline:

Intervention: 147

Control: 177

Important prognostic factors²:

For example

age ± SD:

I: 75.6 ± 5.2

C: 71.5 ± 5.4

Sex:

I: 18% M

C: 12% M

Groups comparable at baseline? Patients in the I group were older, had more comorbidities and more frequent preoperative alcohol use, less often anxiety, less often hearing aid use, greater proportion of pelvic fixation, and a shorter mean aesthetic duration

Describe intervention (treatment/procedure/test):

Once referred to the program patients underwent a pre-operative assessment with a geriatrician within 30 days of their procedure. The geriatricians performed a comprehensive assessment including discussion of goals and expectations for surgery, functional assessment, assessment of nutritional status, evaluation of medical comorbidities, social support, cognitive assessment, evaluation for depression, thorough medication review, and discontinuation of high risk medications when possible. Patients and families also received specific education on delirium and delirium prevention. Results and recommendations from this evaluation
were communicated to the anaesthesia and surgical teams.
Approximately 1–2 weeks prior to surgery, patients attended a presurgical evaluation in the anaesthesiology clinic. American Society of Anesthesiologist (ASA) score,
risk factors, comorbidities were assessed. Anaesthesia planning was performed in coordination with geriatric and surgical teams. Preoperative anaesthesia care included pain control measures such as regional blocks when appropriate. Intraoperatively anaesthesia providers had the option of using a geriatric specific protocol that included avoidance of high-risk aesthetic agents, appropriate pain control, judicious use of muscle relaxants, and maintenance of euvolemia and normothermia.
Postoperatively, patients were co-managed by the primary surgical and geriatric consult team daily until discharge. The geriatric team assisted with adherence to
nonpharmacologic delirium prevention strategies, management of medical comorbidities, pain management, optimizing bowel and bladder function, nutritional status, avoidance of high-risk medications, administration of a daily delirium screen, and facilitating smooth transition to post-acute care.

Additionally, patients in the IG received preoperative geriatric assessment, intraoperative specialized anaesthesia protocol and minimize time under anaesthesia, and postoperative delirium prevention

Describe control (treatment/procedure/test):

Standard care

Length of follow-up:

90 days post-operative

Loss-to-follow-up:

NR

Incomplete outcome data:

NR

Outcome measures and effect size (include 95%CI and p-value if available):

Delirium incidence and recognition

No difference in delirium incidence (I 17 (11.6%) vs C 34 (19.2%); RR = 0.60; CI = 0.35–1.03; NNT = 14; P = .065).

Recognition of patients with delirium significantly increased in the I group (I 13 (76.5%) vs C 8 (23.5%); RR = 3.07; CI = 1.57–6.00; NNT = 3; P = .001*).

Patient outcomes

Lower hospital LOS in the I group (I 4.7 d ± 2.1 vs C 5.4 ± 3.1, p=0.016*).

No difference in ICU LOS (I 0.7 ± 1.0 vs C 0.8 ± 1.3, p=0.40).

No difference in postoperative day walking (I 1.7 ± 1.7 vs C 1.7 ± 1.2, p=0.77).

No difference in 7-day readmission (I 3 (2.1%) vs C 4 (2.3%), p>0.99), 30-day readmission (I 8 (5.4%) vs C 10 (5.7%), p=0.94), and 90-day readmission (I 12 (9.5%) vs C 13 (7.3%), p=0.48).

No comments

Shahrokni, 2020

Type of study:

Retrospective cohort study

Setting and country:

Tertiary care (single-centre) Memorial Sloan Kettering Cancer Centre, New York, United States of America

Funding and conflicts of interests:

Reported

Inclusion criteria:

Patients aged 75 years and older who underwent cancer-related surgical treatment of various cancer types

Exclusion criteria:

Procedures with length of stay of 0 days

Missing mortality data (n = 78)

N total at baseline:

Intervention: 1020

Control: 872

Important prognostic factors

age ± SD:

I: 81 ± 4

C: 80 ± 4

Sex

I: 47.8% M

C: 51.6% M

Groups comparable at baseline?

Intervention group is slightly older. ASA score was similar. Groups were not similar in procedure type (ie. colorectal, gynaecological, urology, thoracic etc.)

Describe intervention:

Geriatic comanagement includes 2 phases: preoperative and postoperative care. Referral for preoperative geriatric evaluation is based on the surgery team’s preference and clinical judgment, not using a formal frailty screening tool.

Preoperative evaluation consists of the electronic Rapid Fitness Assessment. After discussion with the surgery and anesthesiology teams, the geriatrics team recommends interventions aimed at optimizing the patient’s status (ie. Consultation with a cardiologists, exercise recommendations, etc.).

Postoperative phase: geriatrics service visits inpatients on a consultative basis on postoperative day 1 and 3, and further follow up if necessary. Assist with management of comorbid conditions and reintroduction of medications. Implement delirium-reducing interventions, assisting with early mobility, prevention of complications, pain management. Plan for follow up with community clinicians is recommended.

Outpatient care is provided by the surgical team only.

Describe control:

Patients whose care is managed by the surgical service, without geriatric comanagement, may undergo preoperative evaluation by a nongeriatrician (ie. Cardiologist or internist). A geriatric consultation may be requested during the postoperative inpatient period.

Length of follow-up

90 days

Incomplete outcome data

Mortality data missing of 78 patients, integrated in exclusion criteria

Outcome measures and effect size (include 95%CI and p-value if available):

Length of stay:

Intervention group had a longer length of stay (median (IQR)): I 5 (3-8) days vs. C 4 (2-7) days; [p < 0.001)

90-day mortality

Patients in the I group were less likely to die within 90 days after surgical treatment (OR 0.43 95% CI 0.28-0.67; p<0.001).

Adverse surgical outcomes within 30 days (major complication, readmission or emergency room visit)

No difference between groups: OR 0.93 95%CI 0.73-1.18); p = 0.54

Yee, 2022

Type of study:

Prospective cohort study

Setting and country:

One acute hospital (Queen Mary Hospital) and two rehabilitation hospitals (Fung Yiu King Hospital and Maclehose Medical Rehabilitation Hospital) in Hong Kong.

Funding and conflicts of interest:

Reported

Inclusion criteria:

Inclusion criteria were age more than or equal to 65; diagnosis of acute (time of injury within 14 days) isolated hip fracture patients from low energy trauma.

Exclusion criteria:

Exclusion criteria included high-energy trauma, pathological fractures, multiple trauma, or old fractures that occurred more than 2 weeks ago.

N total at baseline:

Intervention: 207

Control: 194

Important prognostic factors²:

For example

age ± SD:

I: 83.6 ± 8.2

C: 84.8 ± 7.6

Sex:

I: 28.5% M

C: 26.8% M

Groups comparable at baseline? No, patients in the I group more often lived at home and had less mean postoperative blood transfusions.

Describe intervention (treatment/procedure/test):

The interventional model differed in the addition of a geriatrician during the postoperative phase, who co-managed the patient in both the acute and rehabilitation hospital. An orthogeriatric speciality ward was set up and implemented in Nov 2018 to concentrate resources and manpower. The new system championed collaboration between the orthopaedic surgeons and the geriatricians. In the acute ward, cases were co-managed during combined ward rounds which occurred 3 times per week. The geriatrician actively reviewed all hip fracture patients to allow prompt diagnosis and management of medical complications, optimization of pain control and monitoring of comorbidities.

Additionally, patients in the IG received expedited and streamlined transferral of patients from acute hospital to rehabilitation hospital

Describe control (treatment/procedure/test):

In the conventional model, patients were admitted to an acute orthopaedic ward where an orthopaedic surgeon was responsible for managing care and treatment of all medical problems, including both orthopaedic and non-orthopaedic–related complications. Medical and geriatric input was only available upon ad hoc request and these interdisciplinary referrals occurred after the complications had already arisen. For semi-urgent referrals, the patients would be reviewed by the physician after 2 days, leading to delays in management and prolonged length of stay.

Length of follow-up:

One year after the index fracture

Loss-to-follow-up:

NR

Incomplete outcome data:

NR

Outcome measures and effect size (include 95%CI and p-value if available):

Length of stay

Less patients with extended LOS in the I group (I 39.1% vs C 64.4%, difference 25.3% (95%CI 15.838% ; 34.767%), p<0.001*).

Lower median acute hospital LOS (I median 7.0 IQR 3-11 days vs C median 8.0 IQR 4-12 days; log-rank P=.001*; unadjusted HR for hospital discharge, 1.284 [95% CI, 1.053–1.566]; P = .01*; adjusted HR, 1.309 [95% CI,1.070–1.602]; P=.009*).

Lower median rehabilitation hospital LOS (I median 16.0 IQR 9-23 vs C median 18 IQR 9-27 days; og-rank P=.001*; unadjusted HR for hospital discharge, 1.391 [95% CI, 1.124–1.723]; P=.002*; adjusted HR, 1.357 [95% CI, 1.095–1.682]; P=.005*).

No difference in preoperative LOS (both median 1 IQR 0-2 dats, log-rank P=0.38).

Lower median postoperative LOS in the I group (I median 4.0 IQR 1-7 vs C median 6.0 IQR 2-10, log-rank p<0.001*).

Mortality

No difference in 30-day mortality (I 1.9% vs C 0%; difference, 1.9% [95% CI, .0%–3.8%]; P=.99), 3-month mortality (I 2.9% vs C 3.1%; difference, 0.2% [95% CI,-3.2% to 3.6%]; P=.56), 6-month mortality (I 4.8% vs C 7.2%; difference, 2.4% [95% CI,-2.4%–7.3%]; P=.24), and 12-month mortality (I 10.6% vs C 13.9%; difference, 3.3% [95% CI,-3.1%–9.7%]; P=.24) after adjusting for covariates.

Functional recovery

Higher median MBI before discharge in the I group (I median 81 IQR 27 vs C median 63.5 IQR 28, p<0.001*).

Higher EMS before discharge from the rehabilitation hospital in the I group (I median 12 IQR 8 vs C 9 IQR 8, p=0.07*).

Complications

Less chest infections in the I group (I 11 (5.3%) vs C 21 (10.8%), difference 5.5% [95%CI 0.2%–10.9%], P=0.04*).

No statistically significant difference in other complications or presence of any medical complications; surgical complications, postoperative blood transfusion, wound complications, urinary tract infection, acute retention of urine, delirium, gastrointestinal bleeding, renal failure, and any medical complications.

Osteoporosis

More prescription of bisphosphonate (I 138 (66.7%) vs C 25 (12.9%); difference, 53.8% [95% CI, 45.8%–61.7%], P<.001*).

There was no difference in the number of subsequent fractures within 1 year of index fracture between the orthogeriatric group and conventional group (3 (1.4%) vs 6 (3.1%), difference, 1.6% [95% CI,-1.3% to 4.6%], P=.27).

Discharge destination from the rehabilitation hospital

No difference in the amount of patients being able to go back to their home (I 96 (69.1%) vs 76 (71.0%); difference, 2.0% [95% CI,-9.6 to 13.5]; P=.74).

Readmission rates

No difference in 28-days hospital unplanned readmission rate (I 12.6% vs C 14.9%, p=0.55).

No difference in readmission due to medical reasons (I 8.2% vs C 11.3%, p=0.55).

No difference in readmission due to orthopaedic reasons (I 4.3% vs C 3.6%, p=0.55).

No comments

Outcome

Result

Heterogeneity

References

Living at home independently

RR 1.03 [0.81; 1.30]

61%

Bub (2022); McDonald (2018); Yee (2022)

Complications

RR 0.74 [0.56; 0.99]

89%

Bub (2022); Fan (2021); Hafner (2021); Kalmet (2019); McDonald (2018); Yee (2022)

Delirium

RR 0.96 [0.53; 1.75]

88%

Fan (2021); Hafner (2021); Kalmet (2019); McDonald (2018); Natesan (2022); Pernik (2021); Yee (2022)

Urinary tract infections

RR 0.55 [0.31; 0.99]

72%

Bub (2022); Fan (2021); Natesan (2022); Yee (2022)

Pneumonia

RR 0.60 [0.42; 0.85]

0%

Bub (2022); Fan (2021); Giannotti (2022) Natesan (2022)

90-day readmission

RR 0.86 [0.43; 1.73]

26%

Giannotti (2022); Pernik (2021)

30-days readmission

RR 0.63 [0.48; 0.83]

0%

Giannotti (2022); McDonald (2018); Natesan (2022); Pernik (2021)

30-day mortality

RR 0.85 [0.53; 1.38]

0%

Fan (2022); Giannotti (2022); Kalmet (2019); Yee (2022)

In-hospital mortality

RR 0.91 [0.16; 5.27]

45%

Bub (2022); Fan (2021); Hafner (2021)

Length of stay

SMD -0.15 [-0.39; 0.09]

85%

Bub (2022); Fan (2021); Hafner (2021); Kalmet (2019); Pernik (2021)

Author (first year)

Exclusion reason

Adogwa (2018)

Wrong population

Aletto (2020)

Wrong population

Anighoro (2020)

Wrong intervention

Baroni (2019)

Wrong population

Baxter (2021)

Full text not available

Blauth (2021)

Wrong population

Blood (2019)

Wrong population

Branas (2018)

Full text not available

Breda (2022)

Wrong population and wrong intervention

Burton (2020)

Wrong intervention

Bryant (2019)

Wrong intervention

Castellví Valls (2018)

Wrong intervention

Chacón-Valenzuela (2022)

Article in Spanish

Cheng (2020)

Wrong population

Christiano (2021)

Wrong intervention

Civinini (2019)

Wrong population

Cooper (2020)

Background article

De Sire (2021)

Wrong population

De Vincentis (2021)

Wrong population and wrong publication type

Deeken (2022)

Wrong population

Deschodt (2020)

Wrong population and wrong intervention

Dottorini (2020)

Background article

Eamer (2018)

Wrong intervention

Festen (2021)

Wrong intervention

Figveld (2019)

Background article

Filippova (2019)

Wrong population

Flikweert (2021)

Wrong population

Fu (2022)

Wrong intervention

Fu (2022)

Wrong intervention

González-Quevedo (2020)

Wrong intervention

González-Quevedo (2022)

Wrong intervention

González-Senac (2021)

Wrong intervention

Greenstein (2019)

Wrong population

Gupta (2021)

Wrong intervention

Handoll (2021)

Wrong population

Igwe (2020)

Review; reference checking performed but no relevant articles

Janssen (2019)

Review; reference checking performed but no relevant articles

Jiménez (2019)

Wrong comparison

Jones (2020)

Background article

Ke (2019)

Background article

Khan (2020)

Wrong population

Kotsani (2021)

Wrong publication type

Kusen (2019)

Wrong population

Kusen (2021)

Wrong population

Kusen (2021)

Wrong population

Laubach (2021)

Wrong population

Lee (2021)

Wrong population

Leung (2018)

Wrong population

Li (2022)

Wrong population

Lin (2021)

Wrong intervention

Liu (2022)

Wrong population

Liu (2022)

Wrong population

Losh (2019)

Wrong population

Mason (2018)

Wrong comparison

McMillan (2022)

Full text not available

Min (2021)

Background article

Mudge (2020)

Wrong population

Mukherjee (2020)

Wrong publication type and wrong population

Murphy (2019)

Wrong population

Oberai (2018)

Wrong population

Ogawa (2022)

Wrong population

Olotu (2019)

Review; reference checking performed but no relevant articles

O'Mara-Gardner (2020)

Wrong intervention

Park (2022)

Wrong population

Parks (2022)

Wrong publication type

Partridge (2018)

Wrong publication type

Patel (2020)

Wrong population

Peng (2020)

Wrong population

Pfeufer (2020)

Wrong comparison

Piolo (2018)

Background article

Portelli Tremont (2022)

Wrong comparison

Quaranta (2021)

Wrong outcomes

Reguant (2019)

Wrong population

Rincón Gómez (2020)

Article in Spanish

Roberts (2021)

Full text not available

Roberts (2022)

Wrong population

Roll (2019)

Wrong intervention

Rostoft (2020)

Wrong comparison

Rubin (2018)

Wrong intervention

Saber (2022)

Wrong population

Sacks (2020)

Background article

Sanli (2019)

Wrong comparison

Schoeneberg (2021)

Wrong population

Schulz (2021)

Wrong population

Sedlock (2020)

Wrong population

Shen (2022)

Wrong study design

Shigemoto (2019)

Wrong intervention

Shigemoto (2021)

Wrong intervention

Shipway (2018)

Wrong population

Steffensmeier (2022)

Wrong population

Stephens (2019)

Wrong population

Styan (2018)

Wrong population

Su (2022)

Wrong population

Subramaniam (2020)

Review; reference checking performed but no relevant articles

Sura-Amonrattana (2021)

Wrong population

Tarazona-Santabalbina (2019)

Wrong population

Tarazona-Santabalbina (2021)

Wrong population

Thillainadesan (2021)

Background article

Thillainadesan (2022)

Review; reference checking performed but no relevant articles

Thillainadesan (2022)

Wrong population: not all patients underwent surgery

Van Dartel (2021)

Wrong population

Van der Vlies (2020)

Wrong intervention

Van Grootven (2021)

Wrong population

Van Heghe (2022)

Wrong population

Vilches-Moraga (2018)

Wrong publication type and wrong population

Wallace (2019)

Wrong population

Werner (2020)

Wrong population

Wiedl (2022)

Wrong population: not all patients underwent surgery

Wu (2019)

Wrong population

Zhang (2022)

Wrong population

Zhao (2022)

Wrong population