29
Missing Data in NSQIP Albumin Peter Doris SQAN

Missing Data in NSQIP

  • Upload
    nantai

  • View
    40

  • Download
    1

Embed Size (px)

DESCRIPTION

Missing Data in NSQIP. Albumin. Peter Doris SQAN. Clarification. Albumen Egg white Albumin A water soluble protein. Why Collect All This Data?. Just report our outcomes We’ll review and decide what to do Literature review Compare with other hospitals - PowerPoint PPT Presentation

Citation preview

Page 1: Missing Data in NSQIP

Missing Data in NSQIP

AlbuminPeter DorisSQAN

Page 2: Missing Data in NSQIP

Clarification

• Albumen– Egg white

• Albumin– A water soluble protein

Page 3: Missing Data in NSQIP

Why Collect All This Data?

• Just report our outcomes• We’ll review and decide what to do

– Literature review– Compare with other hospitals

• Woops! There's no one just like us

• The data is there to calculate “Predicted Risk”– Based on the model, predict “expected outcomes”

• Then we can calculate where we are better or worse

Page 4: Missing Data in NSQIP

The NSQIP Model

• Extensively validated– VA-NSQIP– ACS-NSQIP

• Semi-annual updates

• “Key Variables” identified– Albumin consistently near the top

• I’m convinced – I’ll submit a complete dataset

Page 5: Missing Data in NSQIP

Is Data Missing?

Page 6: Missing Data in NSQIP

Do Lab Values Matter?

Page 7: Missing Data in NSQIP

Labs Often Missing

Page 8: Missing Data in NSQIP

Does Missing Data Matter?

Page 9: Missing Data in NSQIP

What’s Missing? Why?

Page 10: Missing Data in NSQIP

What’s Missing? Why?

Page 11: Missing Data in NSQIP

NSQIP and Missing Values

• Missing data is imputed• Method of imputation is debateable• Does it matter?

Page 12: Missing Data in NSQIP

Missing Data

Page 13: Missing Data in NSQIP
Page 14: Missing Data in NSQIP

Missing Values – Does it Matter?

Page 15: Missing Data in NSQIP

Missing Data

Missing Data in the American College of Surgeons

National Surgical Quality Improvement Program

Are Not Missing at Random: Implications and

Potential Impact on Quality Assessments

Barton H Hamilton, PhD, Clifford Y Ko, MD, MS, MSHS, FACS, Karen Richards, BA,

Bruce Lee Hall, MD, PhD, MBA, FACS

Page 16: Missing Data in NSQIP

Models for Albumin

Page 17: Missing Data in NSQIP

Models for Albumin

Page 18: Missing Data in NSQIP

Albumin

• When measured is a clinical indicator of risk• When not measured the model imputes decreased

risk• When measured in low risk patients the model

assigns higher risk

Page 19: Missing Data in NSQIP

SMH Data2006-2012

n = 6985 completed and transmitted casesMissing Albumin 4022 (57.58%)

With Albumin 2963 (42.42%)

Distribution of Albumin Levels at SMH 2006-2012

Page 20: Missing Data in NSQIP

SMH Data

0.5

11

.5

0 2 4 6 0 2 4 6

Alive Death w/in 30 days

Den

sity

AlbuminGraphs by Death w/in 30 Days

Comparison of Albumin Levels Between Alive and Dead at SMH 2006-2012

There is significant difference between albumin levels between alive and death (P<0.001)

Page 21: Missing Data in NSQIP

SMH Data

• Mortality increases for every 1.7 reduction in albumin level

• Just having your albumin measured increases your odds of death by 4.7

Page 22: Missing Data in NSQIP

Continuum of Care

FIPPA

PIPA PIPA

Page 23: Missing Data in NSQIP

PIPA

The core principle of PIPA that is relevant to physicians is that personal information should not be collected, used, or disclosed without the prior knowledge and consent of the patient, which may be implicit. This principle is subject to limited exceptions. For example:

• Where the collection, use, and/or disclosure is clearly in the interests of the individual and consent cannot be obtained in a timely way.

Page 24: Missing Data in NSQIP

PIPA

Under PIPA, the consent for collection, use, and disclosure of personal information for direct health care

purposes in BC operates primarily on an “implied consent” model. This means that those individuals who

form part of a patient’s “circle of care” (e.g., specialists, referring physicians, lab technologists) can

access, use, disclose, and retain patient information for the purposes of ongoing care and treatment.

Page 25: Missing Data in NSQIP

PIPA and FIPPA

Comparing PIPA and FIPPA

There are some notable differences between PIPA and FIPPA:

PIPA does not include the FIPPA provisions regarding storage and access to personal information from outside Canada. As long as privacy is contractually protected, it does not matter where the data is or where it is accessed from.

PIPA excludes business contact information from the definition of personal information.

PIPA requires consent for the collection, use, and disclosure of personal information. It is up to the organization to determine whether the form of consent is expressed (written or verbal) or deemed (implied).

FIPPA does not contain consent requirements; instead it operates on the principle of “notification” for collection of information.

Page 26: Missing Data in NSQIP

PIPA

Individuals own their personal health information, and physicians act as data stewards, which means that

physicians are responsible and accountable for the personal information they collect, use, and disclose.

Page 27: Missing Data in NSQIP

Can Physicians Disclose Data to NSQIP?

• There’s no implied consent

• Most will cooperate as this is a QI Protocol

• Some will want legal justification

Page 28: Missing Data in NSQIP

PIPA

Roles-based access has great potential to strengthen the trust of patients by ensuring appropriate access to the patient record. However, roles-based access needs also to integrate a “need to know” principle, based on a legitimate relationship with the patient. Unfortunately, “need to know” often becomes “want to know”, so it is necessary to always consider the degree to which access to the personal information istruly needed to perform a given role’s duties.

Page 29: Missing Data in NSQIP

Authority to AccessMedical staff of a hospital

4 (1) A hospital's board must organize a medical staff of which every practitioner regularly practising in the hospital must be a member.

(2) A hospital's board must, after consultation with the executive body of the hospital's medical staff, promulgate bylaws for that medical staff.

(3) A hospital's medical staff must comply with all of the following:

(a) meet regularly in accordance with the Standards issued by the Canadian Council on Health Agency Accreditation;

(b) keep proper minutes of its meetings;

(c) act in an advisory capacity to the hospital's board, in the manner provided in this regulation and provided in the bylaws and rules, if any, of the hospital;

(d) make recommendations regarding the various categories of medical staff membership to be established by the hospital's board and the duties, responsibilities and privileges to be assigned to each category;

(e) assist the hospital's board in providing adequate documentation for the purpose of maintaining a health record for each patient;

(f) participate in appropriate quality improvement activities, including, without limitation, reviewing the following:

(i) deaths occurring in the hospital;

(ii) statistics regarding the progress of patients in the hospital;

(iii) methods of treatment of patients in the hospital;

(iv) results of surgery performed in the hospital;(v) a case when a patient's stay in the hospital is abnormally long;