Senior Rounds – Susan Teschke Can I use venous blood as a substitute for arterial when obtaining blood gases?

Obj ti Objectives 1 List the general indications for blood gas 1. analysis and the complications 2 Critically analyze the usefulness of 2. venous gases in clinical practice 3 Describe 3. D ib situations it ti iin which hi h VBG VBGs might i ht be a good substitution for ABGs

O tli Outline Case Background Information I f ti sought ht from f blood bl d gases Indications

Alternatives Alt ti / Why Wh substitute? b tit t ? Two and a half papers Conclusion Back to the case

Indications/Uses of blood gases Diagnostic criteria ARDS/ALI Respiratory failure Carbon monoxide poisoning Methemoglobinemia

Indications/Uses of blood gases Cutoff values fulfill management recommendations or guidelines Long-term oxygen therapy Preoperative evaluation Pulmonary disability Qualification for noninvasive ventilation Addition of prednisone in PCP Hyperbaric chamber in patients with carbon monoxide poisoning

Indications/Uses of blood gases Assist in the ongoing management Exacerbations of COPD and asthma Course of interstitial lung disease and response to treatment Wean mechanical ventilation, optimize oxygen delivery, limit the risk of oxygen toxicity Assessment of acid-base balance Pathogenesis of disturbance (ex salicylate, toxic OH)

Indications/Uses of blood gases Calculation physiologic variables such as: Oxygen content Oxygen delivery Alveolar-arterial oxygen difference Shunt ratio

Potential complications of an ABG          

Bruising g ((30%)) Pain or tenderness (10%) Hematoma (5%) D Decrease iin pulse l d due tto arterial t i l spasm (1 (1.2% 2% tto 1 1.4%) 4%) Aneurysm (0.2%) Hemorrhage Ischemia Compression neuropathy Sepsis Arteriovenous fistula

STAT!Ref - ACP PIER & AHFS DI® Essentials; Arterial Blood Gases, accessed July 2009

Why use alternative? Avoid complications and pain Avoid invasive monitoring techniques such as indwelling catheter Reduce the number of ABG sampling Arterial access is not available Venous blood draw routinely y and easily y performed

Alternatives Pulse oximetry -> > oxygenation Capnography -> end-tidal PaCO2 Capillary C ill bl blood d gas sampling li Venous blood sampling

FYI  For most of the general indications there are no prospective studies to determine whether ABG analysis alters diagnosis or management or optimizes quality of patient care (11).  One study found that about 30% of arterial blood gases obtained in a tertiary care hospital setting may be unnecessary (12) (12).

STAT!Ref - ACP PIER & AHFS DI® Essentials; Arterial Blood Gases, accessed July 2009

The Question Can I use venous blood as a substitute for arterial when obtaining blood gases?

Case 1 86F with end-stage end stage COPD COPD, admitted with COPD exacerbation and now likely going to require home O2 She has undergone multiple ABGs and has significant bruising and small hematomas at the radial sites Can C we use a venous gas tto quantify tif her h oxygenation?

Case 2 86F with end-stage end stage COPD, admitted with COPD exacerbation and now found to have  LOC Multiple attempts at an ABG are unsuccessful She had a VBG completed in the ER and y yyesterday y as well accidentally Can we use a venous gas to rule out CO2 narcosis as a cause of her  LOC?

Case 3 28M with T1DM seen in ER for suspected DKA Has had multiple episodes in past past, and had arterial spasm with last ABG Adamantly Ad tl refuses f an ABG Can we use a VBG to accurately quantify and document his acidosis and pH?

Rang et al, 2002  Aims: 1. prospectively examine correlations and differences between ABG and peripheral VBG in ED pts 2. Determine whether venous values predict p arterial values accurately enough to replace them in a clinical setting

Rang et al, 2002  Previous studies in ED patients had shown good correlation but..  Retrospective  No agreement on clinically significant differences (± 2 SD?)

Methods  Kingston General Hospital (academic, tertiary care)  Inclusion: all adult pts presenting to ER  Exclusion: 0400-0800h  Data:  Indication  Site, Site time time, number of attempts attempts, age age, sex sex, VS VS, O2sat, meds, O2Rx, presenting complaint, discharge diagnosis

Methods  Venous drawn from same location  Time between  Only ABG results reported

 Identified ALL ABGs completed within a random 6 6-day day period and compared these patients to enrolled patients

Methods  Defining clinically important differences:  Surveyed 45 ERPs  “II would feel uncomfortable using only the venous value for clinical decisions if it was y from the arterial more than ______ units away value.”  pH, pCO2, pO2

Analysis  No primary endpoint defined  Means and CI for pH, pCO2, pO2, survey results  Graphical method to examine agreement  Student’s t-test between means of enrolled vs non-enrolled

Table 1

Rang ett al, R l 2002 T Table bl 1 N=218 enrolled vs n=62 not enrolled No major differences, except slightly more respiratory indication in enrolled g group p vs metabolic in non-enrolled

Results 87% had venous within 5 minutes 96% within 10 minutes 58% (26) surveys returned

Correlation in pH

Rang et al, 2002 graph venous pH to arterial pH, r=0.913

Correlation in pCO2

Rang et al, 2002 graph venous pCO2 to arterial pCO2, r=0.921

Correlation in HCO3

Rang et al, 2002 graph venous HC03 to arterial HC03, 0.953

Agreement in pH – 66%

Rang et al, 2002 graph arterial-venous pH differences to average of arterial and venous (agreement) pH

Agreement in pCO2 – 51%

Rang et al, 2002 graph arterial-venous pCO2 differences to average of arterial and venous (agreement) pCO2

Agreement in HCO3 – 87%

Rang et al, 2002 graph arterial-venous pHCO3 differences to average of arterial and venous (agreement) pHCO3

Arterial-Venous Arterial Venous differences (Mean units))

95% of diff ((2 SD))

pH

Clinically significant g differences ± 0.05

pCO2

± 6.6

± 13.9

HCO3

± 3.5

± 3.5

± 0.08

Conclusions Excellent correlation (linear) Moderate agreement Clinically Cli i ll significant i ifi t diff differences ((survey)) Systematic differences

Therefore… Th f arterial t i l and d venous samples l are not equivalent… BUT Useful to trend (good correlation) Can we accurately calculate arterial values?

Strengths/Weaknesses Small, one center Prospective, checked for bias in enrollment, Canadian Good, simple question, attempted to define clinically significant differences Subjective definition of clinically significant (could we measure this? real time decision? scenarios?) No comment on pO2 No

Toftegaard et al, 2008 and 2009 Several studies correlating pH (a-v) (a v), two studies correlating peripheral venous to arterial pCO2 showing good correlation Few studies correlating pO2, samples not taken simultaneously

Toftegaard et al, 2008 and 2009  Aim: 1. Examine the correlation between arterial values and venous values sampled peripherally, centrally and from a pulmonary arterial catheter, in adult patients with a wide range of acid–base and oxygenation status 2. Evaluate method for calculating arterial pH, pCO2, CO2 HCO3 HCO3, and d pO2 O2 values l ffrom venous samples

Methods  112 adult patients (39♀ (39♀, ages 26-81) 26 81)  From respirology, thoracic ICU, multidisciplinary ICU  Three groups 1. 36 hemodynamically stable patients with COPD 2 51 stable 2. t bl patients ti t without ith t COPD 3. 25 unstable patients without COPD

Methods  Simultaneous arterial and venous gases and SpO2 taken (within 10 seconds)  103 p peripheral, p , 73 central,, 18 mixed venous

 Second arterial taken after venous if p venous sites sampled p multiple  9 patients excluded:  3 missed arterial or venous values  5 had large Hb differences  1 had O2 changed during draw

Methods Peripheral venous drawn from: Upper extremity Well perfused with normal temperature and capillary response No apparent lesions

Anaerobic syringes Compared 1) arterial to venous values and 2) arterial to calculated arterial values

Method

Crazy diagram showing derivation of equation

Results – arterial to venous  No difference before + after arterial samples  Good arterial to peripheral venous correlation for pH ((within acceptable p p lab error))  0.026 ± 0.046 (bias ± 2 SD)

 Moderate a-v correlations for p pCO2 ((outside lab error of ± 5.0, within clinically acceptable error?) – 4.5 ± 8.6 mmHg (bias ± 2 SD)

 Poor a-v correlations for pO2

Results – arterial to calculated arterial Excellent agreement in measured and calculated arterial pH and CO2 pH 0 0.002 002 ± 0.027 0 027 (bias ± 2 SD; lab ± 0.04) 0 04) pCO2 -0.30 ± 3.88 mmHg (bias ± 2 SD; lab ± 5) Similar to variability between consecutive arterial samples (73 pts) SD half of that of direct venous comparison

Poor agreement for pO2 1 57 ± 13.89 1.57 13 89 mmHg (lab ± 4.5) 4 5) for SpO2