[Skip to Navigation]
Sign In
Figure 1.  Data Flow
Data Flow

Flow of data from eligible patients to analytic data set. VESPA indicates Vulnerable Elders Surgical Pathways and Outcomes Assessment.

Figure 2.  Sensitivity and Specificity of the Vulnerable Elders Surgical Pathways and Outcomes Assessment Score
Sensitivity and Specificity of the Vulnerable Elders Surgical Pathways and Outcomes Assessment Score

A, Sensitivity and specificity of the 14-item basic activities of daily living (BADL) and instrumental activities of daily living (IADL) scale (area under the receiver operating characteristic curve, 0.7647). B, Sensitivity and specificity of the 5-item BADL and IADL scale (area under the receiver operating characteristic curve, 0.7632). Receiver operating characteristic curves were used, and they allow users to select a cutoff score based on desired sensitivity and specificity. The area under the receiver operating characteristic curve reflects the overall accuracy of the scores. There is no loss of accuracy when using the abbreviated vs the full set of functional status items (5 vs 14 items). A cutoff score to identify patients for a less burdensome and inexpensive intervention would appropriately select a population with greater sensitivity; for example, using a score of 4 or higher would capture 93% of patients who eventually have postoperative occurrences, whereas a more risky or expensive intervention would more appropriately select a population with less sensitivity, but with higher risk and greater specificity. For example, a cutoff of 10 or higher has a specificity of 80%, which means that patients who do not meet the cutoff identify 80% of the population of patients who will not have a complication.

Table 1.  VESPA Sample Characteristics
VESPA Sample Characteristics
Table 2.  Univariate Logistic Regression for Any Geriatric and Surgical Complications
Univariate Logistic Regression for Any Geriatric and Surgical Complications
Table 3.  Multivariable Logistic Regression Model for Any Geriatric and Surgical Complication
Multivariable Logistic Regression Model for Any Geriatric and Surgical Complication
1.
Elixhauser  A, Andrews  RM.  Profile of inpatient operating room procedures in US hospitals in 2007.  Arch Surg. 2010;145(12):1201-1208.PubMedGoogle ScholarCrossref
2.
Etzioni  DA, Liu  JH, Maggard  MA, O’Connell  JB, Ko  CY.  Workload projections for surgical oncology: will we need more surgeons?  Ann Surg Oncol. 2003;10(9):1112-1117.PubMedGoogle ScholarCrossref
3.
Harari  D, Hopper  A, Dhesi  J, Babic-Illman  G, Lockwood  L, Martin  F.  Proactive care of older people undergoing surgery (‘POPS’): designing, embedding, evaluating and funding a comprehensive geriatric assessment service for older elective surgical patients.  Age Ageing. 2007;36(2):190-196.PubMedGoogle ScholarCrossref
4.
Makary  MA, Segev  DL, Pronovost  PJ,  et al.  Frailty as a predictor of surgical outcomes in older patients.  J Am Coll Surg. 2010;210(6):901-908.PubMedGoogle ScholarCrossref
5.
Dasgupta  M, Rolfson  DB, Stolee  P, Borrie  MJ, Speechley  M.  Frailty is associated with postoperative complications in older adults with medical problems.  Arch Gerontol Geriatr. 2009;48(1):78-83.PubMedGoogle ScholarCrossref
6.
Hirth  VA, Eleazer  GP, Dever-Bumba  M.  A step toward solving the geriatrician shortage.  Am J Med. 2008;121(3):247-251.PubMedGoogle ScholarCrossref
7.
Section for Enhancing Geriatric Understanding and Expertise Among Surgical and Medical Specialists (SEGUE), American Geriatrics Society.  Retooling for an aging America: building the healthcare workforce: a white paper regarding implementation of recommendation 4.2 of this Institute of Medicine Report of April 14, 2008, that ‘All licensure, certification and maintenance of certification for healthcare professionals should include demonstration of competence in care of older adults as a criterion.’.  J Am Geriatr Soc. 2011;59(8):1537-1539.PubMedGoogle ScholarCrossref
8.
Malani  PN.  Functional status assessment in the preoperative evaluation of older adults.  JAMA. 2009;302(14):1582-1583.PubMedGoogle ScholarCrossref
9.
Kothari  A, Phillips  S, Bretl  T, Block  K, Weigel  T.  Components of geriatric assessments predict thoracic surgery outcomes.  J Surg Res. 2011;166(1):5-13.PubMedGoogle ScholarCrossref
10.
Pol  RA, van Leeuwen  BL, Visser  L,  et al.  Standardised frailty indicator as predictor for postoperative delirium after vascular surgery: a prospective cohort study.  Eur J Vasc Endovasc Surg. 2011;42(6):824-830.PubMedGoogle ScholarCrossref
11.
Robinson  TN, Wallace  JI, Wu  DS,  et al.  Accumulated frailty characteristics predict postoperative discharge institutionalization in the geriatric patient.  J Am Coll Surg. 2011;213(1):37-42.PubMedGoogle ScholarCrossref
12.
Robinson  TN, Eiseman  B, Wallace  JI,  et al.  Redefining geriatric preoperative assessment using frailty, disability and co-morbidity.  Ann Surg. 2009;250(3):449-455.PubMedGoogle Scholar
13.
Robinson  TN, Wu  DS, Stiegmann  GV, Moss  M.  Frailty predicts increased hospital and six-month healthcare cost following colorectal surgery in older adults.  Am J Surg. 2011;202(5):511-514.PubMedGoogle ScholarCrossref
14.
Harrington  MB, Kraft  M, Grande  LJ, Rudolph  JL.  Independent association between preoperative cognitive status and discharge location after cardiac surgery.  Am J Crit Care. 2011;20(2):129-137.PubMedGoogle ScholarCrossref
15.
Cronin  J, Livhits  M, Mercado  C,  et al.  Quality improvement pilot program for vulnerable elderly surgical patients.  Am Surg. 2011;77(10):1305-1308.PubMedGoogle Scholar
16.
American College of Surgeons. ACS NSQIP/AGS best practice guidelines: optimal preoperative assessment of the geriatric surgical patient. https://www.facs.org/~/media/files/quality programs/nsqip/acsnsqipagsgeriatric2012guidelines.ashx. Accessed June 28, 2017.
17.
Hall  DE, Arya  S, Schmid  KK,  et al.  Development and initial validation of the Risk Analysis Index for measuring frailty in surgical populations.  JAMA Surg. 2017;152(2):175-182.PubMedGoogle ScholarCrossref
18.
Katz  S, Ford  AB, Moskowitz  RW, Jackson  BA, Jaffe  MW.  Studies of illness in the aged: the index of ADL: a standardized measure of biological and psychosocial function.  JAMA. 1963;185(12):914-919.PubMedGoogle ScholarCrossref
19.
Lawton  MP, Brody  EM.  Assessment of older people: self-maintaining and instrumental activities of daily living.  Gerontologist. 1969;9(3):179-186.PubMedGoogle ScholarCrossref
20.
Podsiadlo  D, Richardson  S.  The timed ‘Up & Go’: a test of basic functional mobility for frail elderly persons.  J Am Geriatr Soc. 1991;39(2):142-148.PubMedGoogle ScholarCrossref
21.
Kroenke  K, Spitzer  RL, Williams  JB.  The Patient Health Questionnaire-2: validity of a two-item depression screener.  Med Care. 2003;41(11):1284-1292.PubMedGoogle ScholarCrossref
22.
Borson  S, Scanlan  J, Brush  M, Vitaliano  P, Dokmak  A.  The Mini-Cog: a cognitive ‘vital signs’ measure for dementia screening in multi-lingual elderly.  Int J Geriatr Psychiatry. 2000;15(11):1021-1027.PubMedGoogle ScholarCrossref
23.
Campbell  DA  Jr, Englesbe  MJ, Kubus  JJ,  et al.  Accelerating the pace of surgical quality improvement: the power of hospital collaboration.  Arch Surg. 2010;145(10):985-991.PubMedGoogle ScholarCrossref
24.
Inouye  SK, Leo-Summers  L, Zhang  Y, Bogardus  ST  Jr, Leslie  DL, Agostini  JV.  A chart-based method for identification of delirium: validation compared with interviewer ratings using the Confusion Assessment Method.  J Am Geriatr Soc. 2005;53(2):312-318.PubMedGoogle ScholarCrossref
25.
Healthcare Cost and Utilization Project. Clinical classifications software for services and procedures. https://www.hcup-us.ahrq.gov/toolssoftware/ccs_svcsproc/ccssvcproc.jsp. Updated March 30, 2017. Accessed January 4, 2014.
26.
Chow  WB, Rosenthal  RA, Merkow  RP, Ko  CY, Esnaola  NF; American College of Surgeons National Surgical Quality Improvement Program; American Geriatrics Society.  Optimal preoperative assessment of the geriatric surgical patient: a best practices guideline from the American College of Surgeons National Surgical Quality Improvement Program and the American Geriatrics Society.  J Am Coll Surg. 2012;215(4):453-466.PubMedGoogle ScholarCrossref
27.
McCarten  JR, Anderson  P, Kuskowski  MA, McPherson  SE, Borson  S.  Screening for cognitive impairment in an elderly veteran population: acceptability and results using different versions of the Mini-Cog.  J Am Geriatr Soc. 2011;59(2):309-313.PubMedGoogle ScholarCrossref
28.
Charlson  M, Szatrowski  TP, Peterson  J, Gold  J.  Validation of a combined comorbidity index.  J Clin Epidemiol. 1994;47(11):1245-1251.PubMedGoogle ScholarCrossref
29.
Charlson  ME, Pompei  P, Ales  KL, MacKenzie  CR.  A new method of classifying prognostic comorbidity in longitudinal studies: development and validation.  J Chronic Dis. 1987;40(5):373-383.PubMedGoogle ScholarCrossref
30.
Birim  O, Maat  AP, Kappetein  AP, van Meerbeeck  JP, Damhuis  RA, Bogers  AJ.  Validation of the Charlson Comorbidity Index in patients with operated primary non-small cell lung cancer.  Eur J Cardiothorac Surg. 2003;23(1):30-34.PubMedGoogle ScholarCrossref
31.
Suidan  RS, Leitao  MM  Jr, Zivanovic  O,  et al.  Predictive value of the Age-Adjusted Charlson Comorbidity Index on perioperative complications and survival in patients undergoing primary debulking surgery for advanced epithelial ovarian cancer.  Gynecol Oncol. 2015;138(2):246-251.PubMedGoogle ScholarCrossref
32.
Fowler  JE  Jr, Terrell  FL, Renfroe  DL.  Co-morbidities and survival of men with localized prostate cancer treated with surgery or radiation therapy.  J Urol. 1996;156(5):1714-1718.PubMedGoogle ScholarCrossref
33.
Kieszak  SM, Flanders  WD, Kosinski  AS, Shipp  CC, Karp  H.  A comparison of the Charlson Comorbidity Index derived from medical record data and administrative billing data.  J Clin Epidemiol. 1999;52(2):137-142.PubMedGoogle ScholarCrossref
34.
Inouye  SK, Peduzzi  PN, Robison  JT, Hughes  JS, Horwitz  RI, Concato  J.  Importance of functional measures in predicting mortality among older hospitalized patients.  JAMA. 1998;279(15):1187-1193.PubMedGoogle ScholarCrossref
35.
Centers for Medicare & Medicaid Services. PFS relative value files. https://www.cms.gov/medicare/medicare-fee-for-service-payment/physicianfeesched/pfs-relative-value-files.html. Accessed July 20, 2016.
36.
Bilimoria  KY, Liu  Y, Paruch  JL,  et al.  Development and evaluation of the universal ACS NSQIP surgical risk calculator: a decision aid and informed consent tool for patients and surgeons.  J Am Coll Surg. 2013;217(5):833-842.e3.PubMedGoogle ScholarCrossref
37.
Suskind  AM, Walter  LC, Jin  C,  et al.  Impact of frailty on complications in patients undergoing common urological procedures: a study from the American College of Surgeons National Surgical Quality Improvement database.  BJU Int. 2016;117(5):836-842.PubMedGoogle ScholarCrossref
38.
Isik  O, Okkabaz  N, Hammel  J, Remzi  FH, Gorgun  E.  Preoperative functional health status may predict outcomes after elective colorectal surgery for malignancy.  Surg Endosc. 2015;29(5):1051-1056.PubMedGoogle ScholarCrossref
39.
Hung  WW, Ross  JS, Boockvar  KS, Siu  AL.  Recent trends in chronic disease, impairment and disability among older adults in the United States.  BMC Geriatr. 2011;11:47.PubMedGoogle ScholarCrossref
40.
Tinetti  ME, Mendes de Leon  CF, Doucette  JT, Baker  DI.  Fear of falling and fall-related efficacy in relationship to functioning among community-living elders.  J Gerontol. 1994;49(3):M140-M147.PubMedGoogle ScholarCrossref
41.
Thomassen  Ø, Storesund  A, Søfteland  E, Brattebø  G.  The effects of safety checklists in medicine: a systematic review.  Acta Anaesthesiol Scand. 2014;58(1):5-18.PubMedGoogle ScholarCrossref
42.
Siriussawakul  A, Nimmannit  A, Rattana-arpa  S, Chatrattanakulchai  S, Saengtawan  P, Wangdee  A.  Evaluating compliance with institutional preoperative testing guidelines for minimal-risk patients undergoing elective surgery.  Biomed Res Int. 2013;2013:835426.PubMedGoogle ScholarCrossref
Original Investigation
December 2017

Estimating Risk of Postsurgical General and Geriatric Complications Using the VESPA Preoperative Tool

Author Affiliations
  • 1Division of Geriatric and Palliative Medicine, Department of Medicine, University of Michigan, Ann Arbor
  • 2Geriatric Research Education Clinical Center, Veterans Affairs Ann Arbor Healthcare System, Ann Arbor, Michigan
  • 3Department of Surgery, University of California, San Francisco
  • 4Department of Surgery, University of Michigan, Ann Arbor
  • 5Department of Otolaryngology, Massachusetts Eye and Ear Infirmary, Boston
JAMA Surg. 2017;152(12):1126-1133. doi:10.1001/jamasurg.2017.2635
Key Points

Question  Can a short, functional, geriatric assessment scale (<10 minutes) administered by surgical nonphysician staff estimate risk of postsurgical complications, including traditional postoperative occurrences and novel geriatric outcomes, such as delirium and falls?

Findings  In this large cohort study of 736 patients 70 years of age or older, this tool estimated risk of postoperative complications, including difficulties with activities of daily living, inability to manage self-care, and number of comorbidities, with excellent statistical fit.

Meaning  Older patients undergoing elective surgery are at more risk than younger patients of postsurgical complications, but those at higher risk can be efficiently identified for closer monitoring.

Abstract

Importance  As greater numbers of older patients seek elective surgery, one approach to preventing postoperative complications is enhanced assessment of risks during preoperative evaluation.

Objective  To determine whether a geriatric assessment tool can be implemented in a preoperative clinic and can estimate risk of postoperative complications.

Design, Setting, and Participants  In this prospective cohort study, patients 70 years of age or older were assessed in a preoperative clinic for elective surgery from July 9, 2008, to January 5, 2011. Patients were screened using the Vulnerable Elders Surgical Pathways and Outcomes Assessment (VESPA) tool developed for this study. Patients were assessed on 5 preoperative activities of daily living recommended by the American College of Surgeons (bathing, transferring, dressing, shopping, and meals), history of falling or gait impairment, and depressive symptoms (2-item Patient Health Questionnaire). Patients also underwent a brief cognitive examination (Mini-Cog) and gait and balance assessment (Timed Up and Go test). A novel question was also asked as to whether patients expected they could manage themselves alone after discharge. Comorbidities and work-related relative value units (categorized into low, moderate, and high tertiles) were also collected. Multivariable logistic regression was performed to estimate risk of postoperative complications. Sustainability of VESPA over time was also evaluated. Medical record review was performed from December 11, 2012, to October 2, 2015, and data analysis was performed from November 15, 2015, to May 18, 2016.

Main Outcomes and Measures  Postoperative surgical and geriatric complications.

Results  Of the 770 patients evaluated, 736 (384 women and 352 men; mean [SD] age, 77.7 [5.7] years) underwent 740 operative procedures; of these patients, 711 had complete data for multivariable analysis. In our sample, 105 patients (14.3%) reported 1 or more difficulties with the 5 activities of daily living, and 270 of 707 patients (38.2%) foresaw themselves unable to manage self-care alone. A total of 131 of 740 patients had geriatric complications, and 114 of 740 patients had surgical complications; 187 of 740 patients (25.3%) had either geriatric or surgical complications. On multivariable analysis, the number of difficulties with activities of daily living (odds ratio [OR], 1.3; 95% CI, 1.0-1.6), anticipated difficulty with postoperative self-care (OR, 1.6; 95% CI, 1.0-2.2), Charlson Comorbidity score of 2 or more vs less than 2 (OR, 1.5; 95% CI, 1.0-2.3), male sex (OR, 1.6; 95% CI, 1.1-2.3), and work-related relative value units (moderate vs low: OR, 1.9; 95% CI, 1.1-3.3; high vs low: OR, 8.8; 95% CI, 5.3-14.5) were independently associated with postoperative complications (overall model area under the receiver operating characteristic curve, 0.77). With these results, a whole-point VESPA score used alone to estimate risk of complications also demonstrated excellent fit (area under the curve, 0.76).

Conclusions and Relevance  Preoperative assessment of older geriatric patients is feasible in the general preoperative clinic and can help identify patients at higher risk of postoperative complications.

Introduction

With advances in surgical technology and with the aging of the population, surgeons are increasingly operating on older, frail patients with complex medical conditions.1,2 Compared with younger patients, this population is at higher risk for postsurgical complications, prolonged hospitalization, and mortality. Past research of preoperative geriatric assessment has relied on an outpatient geriatric clinic3 or a consulting inpatient geriatrician.4,5 However, owing to the worsening national shortage of geriatricians,6,7 we need to develop efficient ways to integrate geriatric assessment into existing surgical care systems. Improved methods of assessing geriatric risk will facilitate targeted proactive interventions to mitigate risk and improve surgical outcomes. Moreover, inclusion of functional status and physical performance increases the ability to assess risk of postoperative complications beyond that seen with traditional risk assessment.3-5,8-17 Improved preoperative assessment of older patients is a major priority of the American College of Surgeons (ACS) and American Geriatrics Society.16

Our research had 2 aims. First, we measured our ability to implement and sustain the Vulnerable Elders Surgical Pathways and Outcomes Assessment (VESPA) preoperative evaluation. The distinguishing hallmark of this intervention was screening performed by the surgical physician assistants (PAs) who added the VESPA assessments to their already existing workload. Second, we investigated whether the VESPA tool could help surgeons recognize patients at higher risk for both postsurgical complications and geriatric postoperative occurrences, such as delirium.

Methods
Design of the VESPA Tool

The VEPSA tool is a translational, multidomain tool administered in less than 10 minutes by the surgical PA. The original investigators (K.H., E.F., and K.M.D) drafted the VESPA tool (eAppendix 1 in the Supplement) by modifying previously developed geriatric assessments of functional status, cognition, and risk of death to increase feasibility of use in the preoperative clinic led by the PA. We included 6 basic activities of daily living (BADL) items (bathing, dressing, transferring, feeding, grooming, and toileting)18 and 8 instrumental activities of daily living (IADL) items (medication administration, meal preparation, telephone use, transportation, shopping, housekeeping, laundry, and finances).19 We allowed for self-reporting or reporting by proxies as to whether patients had difficulty in accomplishing these activities. In addition, we drafted a novel item to the surgical literature: we asked patients (or their proxies), “Can you manage by yourself for several hours alone after your procedure (outpatient)/after discharge (inpatient)?”

Next, we assessed gait and mobility. We asked whether or not the patient had fallen in the past year. The patients then performed a Timed Up and Go test,20 consisting of a timed stand from a chair position, walk 3 m (10 ft) across the floor, and return to a sitting position. Time to completion was categorized into less than 10 seconds (normal), 10 to 20 seconds (borderline risk), and more than 20 seconds (abnormally slow). The clinician qualitatively rated whether or not the patient’s balance was steady vs unsteady during the Timed Up and Go test.

Third, we modified the 2-item Patient Health Questionnaire to ask respondents to recall 1 month (rather than 2 weeks) of anhedonia and depressive symptoms.21 Instead of the original 4-part response, we used a yes or no answer, reducing the burden of delivering the tool verbally. Last, we administered the Mini-Cog Test, an abbreviated dementia screening test.22

Study Protocol

The VESPA tool was implemented in 2008, administered by PAs (overseen by W.P.) in a preoperative surgery clinic in the University of Michigan Health System. The PAs completed the VESPA tool (eAppendix 1 in the Supplement) on paper and retained the test results for future research. A waiver of consent to review medical records and administrative data was approved by the University of Michigan institutional review board (HUM00020657).

We aimed to perform the VESPA assessment on all patients 70 years of age or older who presented to a preoperative clinic for elective surgery from July 9, 2008, to January 5, 2011. Our preoperative clinic evaluates for otolaryngology and oral maxillofacial, plastics, gastrointestinal, urologic, breast, ophthalmologic, thoracic, neurologic, orthopedic, vascular, hernia repair, and solid-organ transplant surgery.

Medical Record Review

From December 11, 2012, to October 2, 2015, we performed a detailed review of inpatient medical records of all patients who completed the VESPA preoperatively. The goal of the review was to ascertain whether the patient underwent the planned surgery, to measure comorbidity burden, and to capture our outcome of interest: postoperative geriatric or surgical complications. For all patients, we used medical record abstraction materials to collect data on postsurgical complications and comorbidities as defined by the National Surgical Quality Improvement Program (NSQIP).23 We developed new materials (available on request) to abstract geriatric comorbidities and postoperative complications including delirium, pressure ulcer, fall, or malnutrition. Delirium was defined as any documentation of a diagnosis of delirium, signs and symptoms of acute confusion, altered mental status, agitation, or the use of restraints or as-needed antipsychotic medications for treatment of delirium symptoms.24 Pressure ulcers did not include other types of skin ulcers (ie, venous, arterial insufficiency, or skin tears). Fall (including near fall) was defined as inadvertent position change from a higher to lower level. Malnutrition was defined when a the patient had a dietician’s note of inadequate energy or protein intake or was eating less than 25% of meals for more than half of his or her meals for more than 5 days. We also captured unplanned readmission in 30 days and death within 30 days.

Administrative Data Review

To evaluate sustainability over time, we collected all administrative health system data for patients 70 years of age or older seen in the preoperative clinic from July 9, 2008, to January 5, 2011, regardless of VESPA evaluation. Common Procedural Terminology and International Classification of Diseases, Ninth Revision procedure codes determined the class of surgery and whether the procedure was performed within 3 months.25

Measures
VESPA Geriatric Risk Factors

We considered each of the VESPA risk factors individually and as composite measures. The individual BADL and IADL were grouped into a 14-item BADL or IADL difficulty count. In addition, we also calculated a shorter functional scale (a 5-item difficulty count) as recommended by the ACS26: transferring, dressing or bathing, preparing meals, and grocery shopping. In our survey, we counted dressing and bathing as 2 separate items. Next, we considered a composite gait and mobility risk indicator that combined any positive fall screening result, Timed Up and Go result more than 20 seconds, and poor results on the balance examination. Third, we used the Mini-Cog score, with a cutoff score (≤3 of 5) indicating cognitive impairment.11,27

Comorbidity

We applied weights in the Charlson Comorbidity Index,28,29 a factor associated with postoperative complications and mortality,30-32 to conditions collected in the medical record review.33 We used a dichotomous variable that was used previously,34 with a cutoff score of less than 2 vs 2 or more.

Complexity of Surgery

The Common Procedural Terminology codes for each patient’s primary procedure were matched with the number of work-related relative value units (WRVUs) from 2013.35 The WRVU is a continuous measure of the overall risk and resources necessary to perform a surgical procedure, allowing for distinction between extensiveness of procedures. To increase the ease of use in a preoperative setting, we tested the WRVU by tertiles (low, moderate, and high) of risk and complexity.

Outcomes of Interest

We combined outcome variables into the following 3 composite outcomes: (1) any postsurgical complication, including any postoperative occurrence (as defined by the NSQIP, such as acute renal failure, pneumonia, incisional surgical site infection, unplanned intubation, sepsis, or urinary tract infection), death within 30 days, and unplanned readmission in 30 days; (2) any geriatric complication (delirium, pressure ulcer, fall, or malnutrition); and (3) any complication (any surgical or geriatric complication).

Statistical Analysis
Implementation and Sustainability

Statistical analysis was performed from November 15, 2015, to May 18, 2016. Using the administrative data linked to the retained VESPA records, we categorized the health system patients into the following 2 groups: (1) patients evaluated with the VESPA tool and (2) control patients who were 70 years of age or older, seen in the same preoperative clinic during the same 6-month period, and underwent similar procedures in the health system within 90 days but who were not evaluated with the VESPA tool. We measured sustainability of the VESPA tool over time by calculating use of the VESPA tool during 6-month periods (for a total of 5 periods). See eAppendix 2 in the Supplement for more detailed methods.

Outcomes Analysis and Development of Abbreviated VESPA Tool

To identify potentially effective risk factors linked to geriatric complications, surgical complications, and the composite outcome of geriatric and surgical complications, we first used univariate logistic regression to determine the statistical and substantive effect on the risk of geriatric, surgical, and combined outcomes using odds ratios (ORs) of greater than 1.2 and P < .05 as criteria for potential retention, respectively. We then entered the candidate risk factors (for all 3 outcomes) into multivariable logistic regression models estimating the combined outcome of both geriatric and surgical complications, using only cases with complete data. Using the ORs from the multivariable model, we assigned whole-point weights by rounding each OR to the nearest integer. Then, for each patient, we could calculate a composite whole-point VESPA score to estimate the risk of complications. Last, we used the area under the receiver operating characteristic curve, which captures overall sensitivity and specificity and overall model fit, to compare the final multivariable vs VESPA score models and determine potential cutoff scores for clinical use. All statistical analyses were conducted using STATA, version 13 (StataCorp LP).

Results
Implementation and Sustainability

We performed 770 VESPA evaluations during the 2.5-year study period; 736 of these 770 patients underwent a surgical procedure at the University of Michigan Health System. Using University of Michigan Health System administrative data, we identified 4269 patients 70 years of age or older who were evaluated during the 2.5-year study window, of whom 3499 were not assessed with VESPA (Figure 1). Among those 3499 patients, we identified 2357 control patients who underwent surgery similar to the primary procedures performed for patients assessed with the VEPSA tool. The overall screening rate during the 2.5 years was 23.8% (736 of 3093). The screening rate was the highest during the first 6 months, when we performed a VESPA evaluation on 42.3% (247 of 584) of eligible patients. In the following 6 months, our evaluation rate decreased to 35.9% (229 of 638), with a decrease to 15.5% (85 of 550), 15.4% (102 of 662), and 10.9% (73 of 667) in the successive 6-month periods (eAppendix 2 in the Supplement).

Outcome Prediction

The mean (SD) age of the 736 patients in the VESPA sample was 77.7 (5.7) years; 384 patients were women and 352 were men. Only 39 patients (5.3%) had a deficiency in any of the BADL, only 177 (24.1%) had a deficiency in any of the IADL, and 181 (24.6%) had difficulty in performing at least 1 of the original 14 BADLs or IADLs prior to surgery. A total of 105 patients (14.3%) had difficulty performing at least 1 of the 5 items on the short functional status survey proposed by the ACS. More than one-third of patients (270 of 707 [38.2%]) had a self-assessed inability to manage themselves for several hours alone after discharge. Other sample characteristics are displayed in Table 1.

Only 4 of the 736 patients assessed with the VESPA tool had 2 different surgical admissions (ie, an original elective surgery with a preplanned second-stage surgery) that could both be linked to the same preoperative evaluations within the 2.5-year study period (Figure 1). We therefore considered the 740 hospitalizations as independent. The most common procedures were otolaryngology and oral maxillofacial, plastics, and gastrointestinal surgery; nearly 476 of 736 procedures (64.7%) were oncologic. When characterizing complexity of surgery (by using tertiles of WRVUs), patients in the least-complex tertile (WRVU < 9.81) underwent minor procedures such as hernia repair, excision of melanoma or lesion, cystoscopy, and biopsy; those in the middle tertile (WRVU = 9.81-15.71) had major procedures such as parathyroidectomy, partial mastectomy, and laporascopy cholecystectomy; and those in the most complex tertile (WRVU ≥ 15.72) underwent cystectomy, colectomy, pancreatectomy, organ transplants, and pelvic exenterations. A complete listing of procedures within the 3 WRVU tertiles can be found in eAppendix 3 in the Supplement.

We found that 131 of the 740 admissions (17.7%) had geriatric complications, and 114 (15.4%) had surgical complications. A total of 187 of 740 patients (25.3%) had a geriatric or surgical complication. The mean length of stay was 3.7 days (range, 1-50 days) (eAppendix 4 in the Supplement).

In unadjusted logistic regressions of geriatric and surgical complications by each risk factor (Table 2), age was not associated with any type of complication. Men were more likely than women to have geriatric complications (OR, 1.9; 95%, CI 1.3-2.8). The composite 14-item BADL and IADL count outcome was associated with geriatric complications (per item: OR, 1.5; 95% CI, 1.0-2.3), whereas the 5-item ACS functional status variable was associated with surgical complications (OR, 1.2; 95% CI, 1.0-1.5). Both the 14-point composite BADL and IADL count (OR, 1.5; 95% CI, 1.0-2.2) and the 5-point ACS variables (OR, 1.2; 95% CI, 1.0-1.4) were associated with the combined outcome. The novel question of self-assessed inability to manage oneself alone after discharge (yes or no) was strongly associated with both complications (geriatric complications: OR, 1.7; 95% CI, 1.1-2.5; surgical complications: OR, 1.7; 95% CI, 1.1-2.6). The dichotomous Charlson Comorbidity Index variable (combined outcome: OR, 1.9; 95% CI, 1.3-2.7) and the high WRVU group (combined outcome: OR, 7.4; 95% CI, 4.6-11.9) were also associated with the outcomes. The remaining VESPA items, such as depression, Mini-Cog Test results, and results of the gait and mobility screening, were not associated with the outcomes.

On multivariable analysis (711 complete cases), we entered the following statistically significant variables from the univariate analysis: functional status (either the 14-item or 5-item combined BADL and IADL difficulty count), self-assessed inability to manage self alone after discharge, sex, Charlson Comorbidity Index, and WRVU tertiles, all of which continued to be independently associated with the outcome (overall model area under the receiver operating characteristic curve, 0.77). The strongest association was a high level of surgical complexity (by WRVU tertile), which estimated that the odds of a complication would be 7 times higher compared with the least complex tertile (Table 3). The whole-point composite VESPA scores were also associated with the development of either geriatric or surgical complications, with an area under the receiver operating characteristic curve of 0.76 for both the 14-item and 5-item combined BADL and IADL models (Figure 2). By designating a VESPA score of 9 or more as the cutoff for elevated estimated complication risk and comparing the predicted vs the actual complication outcome, we found that sensitivity would be 65% for the VESPA score based on the 14-item functional status score and would be 68% using the 5-item ACS score, while specificity was 76% whether the long or short functional status lists were used.

Discussion

In a large health care system’s preoperative clinic led by PAs, we prospectively performed a geriatric assessment on 736 older patients. Without screening, we would have missed one-fourth of patients with a preoperative functional deficit and 38% with a novel risk factor: those who judged themselves as not being able to manage themselves alone after discharge. Both of these factors were significantly associated with a composite outcome of postoperative complications. Our finding that preoperative functional impairment is associated with postoperative complications is consistent with findings from small, prospective single-center studies, underscoring the need to implement system-wide efforts.3,5,11,36-39 A larger single-center study assessed 4 BADLs (no IADL) preoperatively, but outcomes were limited to those found in the NSQIP (ie, they did not include geriatric outcomes).17

We validated an abbreviated 5-item functional status scale proposed by the ACS, NSQIP, and American Geriatrics Society.26 This scale was associated with complications just like the more time-consuming 14-item functional assessment. The whole-point composite VESPA scale was useful for identifying postoperative risk of complications. Using a cutoff score of 9 or higher, the specificity of the shorter VESPA scale was a respectable 76% and the sensitivity was a more modest 68% for postoperative complications. Therefore, we will be using an abbreviated version of VESPA at our institution (eAppendix 1 in the Supplement). We expect that this more efficient version of the VESPA tool will facilitate sustainability of assessment rates in the future.

To our knowledge, this is the first study incorporating patients’ self-assessed inability to manage themselves after an operation. The basis of the question was grounded in our clinical expertise and in other geriatric self-efficacy literature (eg, confidence in ADL performance without falling).40 Because this single item was associated with outcomes independent of the other risk factors and surgical complexity, we believe that this item captures the patient’s (or caregiver’s) additional insight into capacity beyond what is captured by the other measures. Our results suggest that this item is efficient and effective in the preoperative evaluation of older patients undergoing surgery and deserves further validation.

The perioperative quality improvement literature is extensive,41 but published research on the success of implementing a quality improvement intervention in the preoperative setting within a clinical health system is nearly nonexistent. To our knowledge, this is the first study of an additional preoperative geriatric evaluation to report a compliance rate (42% in the first 6 months, with decreasing rates thereafter). One prior study42 using administrative data to evaluate preoperative quality of standard testing (ie, laboratory tests and radiographs) reported a compliance rate of only 12%. We have no benchmark to judge whether our initial compliance rate was high or low, only that we were successful at implementing the screening in a large number of patients, mostly in the first year of the study. Our results suggest that additional intervention is needed to better sustain compliance after 1 year. Although the VESPA tool is brief and can be performed in less than 10 minutes, the PAs were not allotted additional time for its performance. Naturally, we had staff turnover; however, we did not repeat formal training after initiation of the VESPA tool. We also transitioned to a new electronic health record after the first year, which did not include conversion of the VESPA tool from its paper-only data collection.

Therefore, the next phase will implement VESPA into the electronic health record, including automated reminders for the targeted population based on age. The target population will be streamlined to only patients who are expected to stay at least 1 night for the elective procedure. Second, we will implement annual reinforcement in the tool for existing and new PAs, including quarterly feedback of VESPA evaluation rates and results. Third, we will validate the abbreviated VESPA scale in a new sample of patients. Last, those with a positive VESPA score of 9 or more will be referred for additional interventions (eg, delirium prevention for those with preexisting cognitive impairment and physical therapy for those with gait and mobility impairment). Thus, we plan to expand the VESPA score to a clinical intervention, which will test whether the VESPA tool can improve surgical and geriatric outcomes.

Limitations

Our study has limitations. The highest surgical complexity group had the strongest association with both geriatric and surgical complications but was based on WRVUs and the Common Procedural Terminology codes submitted with the procedure rather than during the preoperative visit. eAppendix 4 in the Supplement is a guide to categorize patients into similar low-, moderate-, and high-complexity groups. We believe that PAs can easily assign patients to 3 categories of surgery complexity. The prospectively calculated VESPA score needs further validation; first within our own center in a new sample of patients and then, ultimately, in other environments, such as centers without preoperative clinics.

Conclusions

A geriatric evaluation that estimates postoperative surgical and geriatric complications can feasibly be implemented into a preoperative clinic and performed by nonsurgeon health care personnel; however, in our study, use of the tool decreased over time. A streamlined tool with only 5 ADL items has great potential for more efficient future implementation.

Back to top
Article Information

Corresponding Author: Kathleen M. Diehl, MD, Department of Surgery, University of Michigan, 3303 Cancer Center, SPC 5392, Ann Arbor, MI 48109 (kdiehl@umich.edu).

Accepted for Publication: April 14, 2017.

Published Online: August 2, 2017. doi:10.1001/jamasurg.2017.2635

Author Contributions: Dr Min and Ms Chan had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Drs Min and Hall shared first authorship.

Study concept and design: Min, Hall, Finlayson, Englesbe, Palazzolo, Diehl.

Acquisition, analysis, or interpretation of data: Min, Hall, Palazzolo, Chan, Hou, Miller, Diehl.

Drafting of the manuscript: Min, Palazzolo, Chan, Diehl.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Min, Chan, Hou, Miller.

Obtained funding: Min, Diehl.

Administrative, technical, or material support: Min, Englesbe, Chan, Diehl.

Study supervision: Min, Hall, Finlayson, Englesbe.

Conflict of Interest Disclosures: Dr Finlayson reported being a Founder’s shareholder in Ooney Inc. No other conflicts were reported.

Funding/Support: This research was supported by grant NIA AG024824 from the University of Michigan Older Americans Independence Act Claude D. Pepper Center. Dr Min was also supported by a Pepper Center Research Career Development Core grant and the Hartford Foundation Center of Excellence.

Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

References
1.
Elixhauser  A, Andrews  RM.  Profile of inpatient operating room procedures in US hospitals in 2007.  Arch Surg. 2010;145(12):1201-1208.PubMedGoogle ScholarCrossref
2.
Etzioni  DA, Liu  JH, Maggard  MA, O’Connell  JB, Ko  CY.  Workload projections for surgical oncology: will we need more surgeons?  Ann Surg Oncol. 2003;10(9):1112-1117.PubMedGoogle ScholarCrossref
3.
Harari  D, Hopper  A, Dhesi  J, Babic-Illman  G, Lockwood  L, Martin  F.  Proactive care of older people undergoing surgery (‘POPS’): designing, embedding, evaluating and funding a comprehensive geriatric assessment service for older elective surgical patients.  Age Ageing. 2007;36(2):190-196.PubMedGoogle ScholarCrossref
4.
Makary  MA, Segev  DL, Pronovost  PJ,  et al.  Frailty as a predictor of surgical outcomes in older patients.  J Am Coll Surg. 2010;210(6):901-908.PubMedGoogle ScholarCrossref
5.
Dasgupta  M, Rolfson  DB, Stolee  P, Borrie  MJ, Speechley  M.  Frailty is associated with postoperative complications in older adults with medical problems.  Arch Gerontol Geriatr. 2009;48(1):78-83.PubMedGoogle ScholarCrossref
6.
Hirth  VA, Eleazer  GP, Dever-Bumba  M.  A step toward solving the geriatrician shortage.  Am J Med. 2008;121(3):247-251.PubMedGoogle ScholarCrossref
7.
Section for Enhancing Geriatric Understanding and Expertise Among Surgical and Medical Specialists (SEGUE), American Geriatrics Society.  Retooling for an aging America: building the healthcare workforce: a white paper regarding implementation of recommendation 4.2 of this Institute of Medicine Report of April 14, 2008, that ‘All licensure, certification and maintenance of certification for healthcare professionals should include demonstration of competence in care of older adults as a criterion.’.  J Am Geriatr Soc. 2011;59(8):1537-1539.PubMedGoogle ScholarCrossref
8.
Malani  PN.  Functional status assessment in the preoperative evaluation of older adults.  JAMA. 2009;302(14):1582-1583.PubMedGoogle ScholarCrossref
9.
Kothari  A, Phillips  S, Bretl  T, Block  K, Weigel  T.  Components of geriatric assessments predict thoracic surgery outcomes.  J Surg Res. 2011;166(1):5-13.PubMedGoogle ScholarCrossref
10.
Pol  RA, van Leeuwen  BL, Visser  L,  et al.  Standardised frailty indicator as predictor for postoperative delirium after vascular surgery: a prospective cohort study.  Eur J Vasc Endovasc Surg. 2011;42(6):824-830.PubMedGoogle ScholarCrossref
11.
Robinson  TN, Wallace  JI, Wu  DS,  et al.  Accumulated frailty characteristics predict postoperative discharge institutionalization in the geriatric patient.  J Am Coll Surg. 2011;213(1):37-42.PubMedGoogle ScholarCrossref
12.
Robinson  TN, Eiseman  B, Wallace  JI,  et al.  Redefining geriatric preoperative assessment using frailty, disability and co-morbidity.  Ann Surg. 2009;250(3):449-455.PubMedGoogle Scholar
13.
Robinson  TN, Wu  DS, Stiegmann  GV, Moss  M.  Frailty predicts increased hospital and six-month healthcare cost following colorectal surgery in older adults.  Am J Surg. 2011;202(5):511-514.PubMedGoogle ScholarCrossref
14.
Harrington  MB, Kraft  M, Grande  LJ, Rudolph  JL.  Independent association between preoperative cognitive status and discharge location after cardiac surgery.  Am J Crit Care. 2011;20(2):129-137.PubMedGoogle ScholarCrossref
15.
Cronin  J, Livhits  M, Mercado  C,  et al.  Quality improvement pilot program for vulnerable elderly surgical patients.  Am Surg. 2011;77(10):1305-1308.PubMedGoogle Scholar
16.
American College of Surgeons. ACS NSQIP/AGS best practice guidelines: optimal preoperative assessment of the geriatric surgical patient. https://www.facs.org/~/media/files/quality programs/nsqip/acsnsqipagsgeriatric2012guidelines.ashx. Accessed June 28, 2017.
17.
Hall  DE, Arya  S, Schmid  KK,  et al.  Development and initial validation of the Risk Analysis Index for measuring frailty in surgical populations.  JAMA Surg. 2017;152(2):175-182.PubMedGoogle ScholarCrossref
18.
Katz  S, Ford  AB, Moskowitz  RW, Jackson  BA, Jaffe  MW.  Studies of illness in the aged: the index of ADL: a standardized measure of biological and psychosocial function.  JAMA. 1963;185(12):914-919.PubMedGoogle ScholarCrossref
19.
Lawton  MP, Brody  EM.  Assessment of older people: self-maintaining and instrumental activities of daily living.  Gerontologist. 1969;9(3):179-186.PubMedGoogle ScholarCrossref
20.
Podsiadlo  D, Richardson  S.  The timed ‘Up & Go’: a test of basic functional mobility for frail elderly persons.  J Am Geriatr Soc. 1991;39(2):142-148.PubMedGoogle ScholarCrossref
21.
Kroenke  K, Spitzer  RL, Williams  JB.  The Patient Health Questionnaire-2: validity of a two-item depression screener.  Med Care. 2003;41(11):1284-1292.PubMedGoogle ScholarCrossref
22.
Borson  S, Scanlan  J, Brush  M, Vitaliano  P, Dokmak  A.  The Mini-Cog: a cognitive ‘vital signs’ measure for dementia screening in multi-lingual elderly.  Int J Geriatr Psychiatry. 2000;15(11):1021-1027.PubMedGoogle ScholarCrossref
23.
Campbell  DA  Jr, Englesbe  MJ, Kubus  JJ,  et al.  Accelerating the pace of surgical quality improvement: the power of hospital collaboration.  Arch Surg. 2010;145(10):985-991.PubMedGoogle ScholarCrossref
24.
Inouye  SK, Leo-Summers  L, Zhang  Y, Bogardus  ST  Jr, Leslie  DL, Agostini  JV.  A chart-based method for identification of delirium: validation compared with interviewer ratings using the Confusion Assessment Method.  J Am Geriatr Soc. 2005;53(2):312-318.PubMedGoogle ScholarCrossref
25.
Healthcare Cost and Utilization Project. Clinical classifications software for services and procedures. https://www.hcup-us.ahrq.gov/toolssoftware/ccs_svcsproc/ccssvcproc.jsp. Updated March 30, 2017. Accessed January 4, 2014.
26.
Chow  WB, Rosenthal  RA, Merkow  RP, Ko  CY, Esnaola  NF; American College of Surgeons National Surgical Quality Improvement Program; American Geriatrics Society.  Optimal preoperative assessment of the geriatric surgical patient: a best practices guideline from the American College of Surgeons National Surgical Quality Improvement Program and the American Geriatrics Society.  J Am Coll Surg. 2012;215(4):453-466.PubMedGoogle ScholarCrossref
27.
McCarten  JR, Anderson  P, Kuskowski  MA, McPherson  SE, Borson  S.  Screening for cognitive impairment in an elderly veteran population: acceptability and results using different versions of the Mini-Cog.  J Am Geriatr Soc. 2011;59(2):309-313.PubMedGoogle ScholarCrossref
28.
Charlson  M, Szatrowski  TP, Peterson  J, Gold  J.  Validation of a combined comorbidity index.  J Clin Epidemiol. 1994;47(11):1245-1251.PubMedGoogle ScholarCrossref
29.
Charlson  ME, Pompei  P, Ales  KL, MacKenzie  CR.  A new method of classifying prognostic comorbidity in longitudinal studies: development and validation.  J Chronic Dis. 1987;40(5):373-383.PubMedGoogle ScholarCrossref
30.
Birim  O, Maat  AP, Kappetein  AP, van Meerbeeck  JP, Damhuis  RA, Bogers  AJ.  Validation of the Charlson Comorbidity Index in patients with operated primary non-small cell lung cancer.  Eur J Cardiothorac Surg. 2003;23(1):30-34.PubMedGoogle ScholarCrossref
31.
Suidan  RS, Leitao  MM  Jr, Zivanovic  O,  et al.  Predictive value of the Age-Adjusted Charlson Comorbidity Index on perioperative complications and survival in patients undergoing primary debulking surgery for advanced epithelial ovarian cancer.  Gynecol Oncol. 2015;138(2):246-251.PubMedGoogle ScholarCrossref
32.
Fowler  JE  Jr, Terrell  FL, Renfroe  DL.  Co-morbidities and survival of men with localized prostate cancer treated with surgery or radiation therapy.  J Urol. 1996;156(5):1714-1718.PubMedGoogle ScholarCrossref
33.
Kieszak  SM, Flanders  WD, Kosinski  AS, Shipp  CC, Karp  H.  A comparison of the Charlson Comorbidity Index derived from medical record data and administrative billing data.  J Clin Epidemiol. 1999;52(2):137-142.PubMedGoogle ScholarCrossref
34.
Inouye  SK, Peduzzi  PN, Robison  JT, Hughes  JS, Horwitz  RI, Concato  J.  Importance of functional measures in predicting mortality among older hospitalized patients.  JAMA. 1998;279(15):1187-1193.PubMedGoogle ScholarCrossref
35.
Centers for Medicare & Medicaid Services. PFS relative value files. https://www.cms.gov/medicare/medicare-fee-for-service-payment/physicianfeesched/pfs-relative-value-files.html. Accessed July 20, 2016.
36.
Bilimoria  KY, Liu  Y, Paruch  JL,  et al.  Development and evaluation of the universal ACS NSQIP surgical risk calculator: a decision aid and informed consent tool for patients and surgeons.  J Am Coll Surg. 2013;217(5):833-842.e3.PubMedGoogle ScholarCrossref
37.
Suskind  AM, Walter  LC, Jin  C,  et al.  Impact of frailty on complications in patients undergoing common urological procedures: a study from the American College of Surgeons National Surgical Quality Improvement database.  BJU Int. 2016;117(5):836-842.PubMedGoogle ScholarCrossref
38.
Isik  O, Okkabaz  N, Hammel  J, Remzi  FH, Gorgun  E.  Preoperative functional health status may predict outcomes after elective colorectal surgery for malignancy.  Surg Endosc. 2015;29(5):1051-1056.PubMedGoogle ScholarCrossref
39.
Hung  WW, Ross  JS, Boockvar  KS, Siu  AL.  Recent trends in chronic disease, impairment and disability among older adults in the United States.  BMC Geriatr. 2011;11:47.PubMedGoogle ScholarCrossref
40.
Tinetti  ME, Mendes de Leon  CF, Doucette  JT, Baker  DI.  Fear of falling and fall-related efficacy in relationship to functioning among community-living elders.  J Gerontol. 1994;49(3):M140-M147.PubMedGoogle ScholarCrossref
41.
Thomassen  Ø, Storesund  A, Søfteland  E, Brattebø  G.  The effects of safety checklists in medicine: a systematic review.  Acta Anaesthesiol Scand. 2014;58(1):5-18.PubMedGoogle ScholarCrossref
42.
Siriussawakul  A, Nimmannit  A, Rattana-arpa  S, Chatrattanakulchai  S, Saengtawan  P, Wangdee  A.  Evaluating compliance with institutional preoperative testing guidelines for minimal-risk patients undergoing elective surgery.  Biomed Res Int. 2013;2013:835426.PubMedGoogle ScholarCrossref
×