Does Preoperative Hemoglobin Value Predict Postoperative Cardiovascular Complications after Total Joint Arthroplasty?

1
Does Preoperative Hemoglobin Value Predict
Postoperative Cardiovascular Complications after
Total Joint Arthroplasty?
Kishor Gandhi MD, MPH, Eugene Viscusi MD, Luis
Pulido MD, and Javad Parvizi, MD
2
Introduction

• Anemia defined by the World Health Organization
  (WHO)1
• Hemoglobin (Hb) lt12 g/dl in women
• Hemoglobin (Hb) lt13 g/dl in men
• The incidence of anemia in the U.S. agegt70 is
  approximately 132.
• Literature shows
• Mild degrees of anemia or polycythemia increased
  risk of 30-day mortality and cardiac events in
  older male veterans undergoing non-cardiac
  surgeries3.
• Multicenter Cohort study found preoperative
  anemia to be independently associated with
  adverse outcomes (death, stroke, or acute kidney
  injury) after cardiac surgery4.
• Preoperative anemia may have independent harmful
  effects due to limited patient physiologic
  reserve and added stress of surgery.

1. WHO. Technical Report Series No. 405, Geneva,
Switzerland WHO.1968. 2. Penninx et al.
Journal of Gerontology. 2006. Vol. 61A.
474-479. 3. Wu et al. JAMA. 2007 297
2481-2488. 4. Karkouti et al. Circulation
2008117478-484.
3
Introduction (contd)

• Limited information is available on the impact of
  preoperative anemia in the elderly population
  following Total Joint Arthroplasty (TJA).
• Total joint replacements are increasing steadily
  each year to an estimated 3.48 million primary
  knee and 572,000 primary hip replacements by the
  year 2030.
• Epidemiologic implications are significant.
• Objective
• Examine the association between preoperative
  hemoglobin levels and cardiovascular
  complications in patients undergoing total hip
  and knee replacements.

4
Methods

• Following IRB approval of this retrospective
  study the sample (n3954) consisted of all
  patients who underwent primary hip and knee
  arthroplasty during the years 2004 and 2005.  
• Patients were followed postoperatively during
  their hospitalization and were observed for
  complications.
• The past medical history of all patients was
  extracted from the preoperative evaluation office
  records and linked to the database containing
  patient information on complications.

5
Analysis

• All statistical analysis was conducted with use
  of SPSS software (version 11.0, Chicago,
  Illinois).
• Bivariate analyses were conducted on the three
  comparison groups categorized by preoperative
  hemoglobin values
• 1. Less than 12 g/dl (anemia group)
• 2. 12-14 g/dl (normal reference group)
• 3. Greater than 14 g/dl (above normal group)
• Primary outcomes measured were the occurrence of
  any postoperative cardiovascular complications
• Angina, arrhythmias, bradycardia, asystole,
  CHF, myocardial Infarctions, compartment
  syndrome, DVTs, hypertension, hypotension,
  hypovolemic shock, and vascular injuries
•  Multivariate analyses controlled for potential
  confounders including
• Age, BMI, race, gender, O.R. time, joint type
  (hips vs. knees), number of joints (unilateral
  vs. bilateral), DM, HTN, cardiac history (CAD,
  arrhythmia, CHF, valvular disorder), PVD, DVT,
  and previous placement of Greenfield filter.

6
ResultsTable 1-Patient Demographics (n3954)
7
Results Table 2 Bivariate analysis of
cardiovascular complications among groups (Total
Sample Size3954)
Complications Hgblt12 (n742) Hgb12-14 (n1914) Hgbgt14 (n1298) P value
Cardiovascular (n116) 41 (5.5) 47 (2.5) 28 (2.2) Plt0.05

• Cardiac complication include Angina, Tachycardia
  (SVTs), Arrhythmias (A-fib), bradycardia,
  Asystole, CHF, and Myocardial Infarction.
• Vascular complications include Compartment
  syndrome, DVTs, Hypertension, Hypotension,
  Hypovolemic shock, and Vascular injury.

8
ResultsTable 3 Logistic Regression Model for
Cardiovascular Complications
9
ResultsTable 4 Logistic Regression Model
Summary
Hemoglobin Beta S.E. P value Odds Ratio
Hgblt12 0.575 0.219 0.022 1.778
Hgb 12-14 Reference Reference Reference Reference
Hgbgt 14 0.024 0.274 0.931 1.024
10
Discussion

• Ø  In this analysis, 742 patients had
  preoperative anemia, 1914 patients with
  hemoglobin levels in the reference range, and
  1298 patients in the above normal group (Table
  1).
• Ø   Average age of patients were 65.9 (Hgblt12),
  65.5 (reference group), and 60.7 years (Hgbgt14).
  
•  
• Ø  The anemia group had longer average length of
  surgery (128.3 minutes), compared to the
  reference and above normal groups (114.3 and
  116.1 minutes, respectively).
•  
• Ø  A higher percentage of patients with
  preoperative anemia had a history of CAD (16.8),
  CHF (3.6), arrhythmias (7), and valvular
  disorders (11).
•  
• Ø Increased occurrence of comorbid hypertension
  (59.6), diabetes (3.6), peripheral vascular
  disease (1.7), and DVTs (3.6) were also noted
  in this group.
•  
• Ø Bivariate analyses indicated that patients with
  preoperative anemia have greater cardiovascular
  complications (Table 2).
• Ø  Due to significant differences in potential
  confounders between the three groups bivariate
  analyses, they were controlled for in logistic
  regression models (Table 3).
• Ø  Logistic regression model showed that patients
  with preoperative anemia (Hgblt12) have
  statistically significant risk of cardiovascular
  complications (Odds Ratio1.778, plt0.05).

11
Conclusion

• Ø   In this retrospective analysis, anemic
  patients (those identified with lower
  preoperative hemoglobin values) have an increased
  risk of cardiovascular complications.
•  
• Ø  Preoperative hemoglobin levels may be
  predictive of postoperative cardiovascular
  complications.
•  
• Ø  These patients may benefit from more than
  just reversal of anemia preoperatively, rather
  increased surveillance for postoperative
  complications.
•  
•  

12
Limitations and Implications

• Limitations of our study include
• Small sample size relative to other published
  studies (JAMA).
• Retrospective analyses may result in omission of
  information regarding co-morbidities experienced
  by patients.
• There is the potential for selection bias because
  highest risk patients may not qualify for TJA.
• Information of intraoperative transfusion was not
  in the database and was not considered in the
  analysis.
• Implications
• Patients identified to be at higher risk of
  cardiovascular risk may benefit from additional
  surveillance postoperatively.