Tag Archives: SAS Programmer

Cracking the Code: Estimate at Completion (EAC) of a Clinical Trial

Project management is a continuous loop of planning what to do, checking on progress, comparing progress to plan, taking corrective action if needed, and re-planning. The fundamental items to plan, monitor, and control are time, cost, and performance so that the project stays on schedule, does not exceed its budget, and meets its specifications.  Of course all of these activities are based on having an agreed upon Work Breakdown Structure (tasks/activities) on which to base the schedule and cost estimates.  During the planning phase of a project, the project manager with the assistance of the project team needs to define the process and procedures that will be used during the implementation phase to monitor and control the project’s performance.

Productivity in the pharmaceutical/biotech/medical device industry is going down. Some compounds have reached the billions expenditures cost without any guarantee that it will ever be approved or reach the market.  So how can we evaluate the performance of some of these clinical trials?

I will not go into details in the degree of project management activities managed and performed by a data manager since this can vary widely per company.  A good clinical data manager or manager of data management should be able to implement basic PM principles that will improve quality and timeliness of a clinical trial, regardless if the trial is fully outsourced (e.g. CRO performed most of the work).

You can find my article about the Role of Project Management in Clinical Data Management (2012) here for further reading.

So what is Estimate at Completion or EAC? or What is the project likely to cost?

There are several methods we could use to calculate EAC.

Let’s look at one formula. EAC =  AC (Actual Cost) + ETC (Estimate to Complete)  so what happens when you don’t know the ETC?

We could use the following formula to derive that value: ETC = (BAC – EV) / CPI =>>>>??? So what? More formulas? How do I get BAC or EV or CPI?

Let’s look at those in more details.

 BAC =>>>Budget at Completion (how much did you
budget for the total project?)
CPI =>>> Cost Performance Index (CPI): BCWP/ACWP

EV = Earned Value

Earned Value Analysis example for a phase 1 trial (*figures in the thousands / millions = fictitious  numbers)

The final clinical trial results includes 100 subjects. The estimated cost is $20 per subject.  That results in an estimated budget of $2000 (100 x 20). During the planning, the CRO indicated that would be able to enroll 5 subjects per week.  Therefore the estimated duration of the trial is 20 weeks (100 / 5)

EV blocks: From the project plan

Estimated Budget: $2000

Estimated Schedule: 20 weeks

Planned Value (PV): at the end of the trial is $2000

Variance between planned and actual at the end of the first week:

Based on the estimated scheduled, I should have 25 subjects enrolled. At $20 per subject, the planned value at the end of the week is $500 (25 x 20)

PV = $500

At the end of the first week, the CRO reports that he has enrolled 20 subjects  and the actual cost of that study is $450. With this information we can look at schedule and cost variance.

SV = EV – PV

SV = $400 – $500 = – 100 ($100 work of subject recruitment is behind schedule).

CV = EV – AC

CV = $400 – $450 = -50 ($50 work of the project is over budget)

*negative figures means bad.

Using early results to predict later results:

Schedule Performance Index (SPI)

SPI = EV/PV

SPI = 400/500 = .80

Cost Performance Index (CPI)

CPI = EV/AC

CPI = 400/450 = .89 –> over budget or expending more

These rations can be used to estimate performance of the project to completion based on the early actual experience.

Estimate to Completion (ETC)
ETC= (PV at completion) – EV)/CPI

ETC= (2000 – 400)/CPI

ETC = (1600/.89) =$ 1798 from end of week one (after 5 days) and it will take additional $1798 to complete the study

Estimate at Completion (EAC)

EAC = AC + ETC

EAC = 450 + 1798 = $2248

If nothing changes, based on the actual results at the end of the first week, the study is estimated  to cost $2248 (rather than the planned cost of $2000) and will take 20 percent longer.

The formulas assumes that the accumulative performance reflected in the CPI is likely to continue for the duration of the project.

You do not need to memorize all of these formulas. There are plenty of tools in the industry that does the computation for you. But if you do not have it available, you can use Excel, set-up your template and plug in the numbers.

Earned Value

 

 

 

 

 

 

 

As per PMI – PMBOK definition, Cost management “…includes the processes involved in estimating, budgeting, and controlling costs so that the project can be completed within the approved budget.”   A Guide to the Project Management Body of Knowledge (PMBOK® Guide).

We have shown you, that PM tools such as Earned Value  Analysis, can be applied to clinical trials or specific work break down (WBS) activities within the data management team.

Based on the above outcome of the project performance related to the schedule, the data manager should be able to determine if she should modify the current plan or revise the original plan.

It is a perfect tool for data managers and managers of data managers and could be part of your risk based processes.

If bringing efficiency, improving data quality and significantly reducing programming time after implementing CDISC standards is on your radar screen, I’d love to chat when it’s convenient. All the best.

Anayansi Van Der Berg has an extensive background in clinical data management as well as experience with different EDC systems including Oracle InForm, InForm Architect, Central Designer, CIS, Clintrial, Medidata Rave, Central Coding, OpenClinica Open Source and Oracle Clinical. SAS, CDASH/SDTM (CDISC standards implementation and mapping), SAS QC checks and clinical data reporting.

Source:

A Guide to the Project Management Body of Knowledge (PMBOK® Guide).

Notes from my PM class at Keller 2007-2009

Images – Google images

FAIR USE-
“Copyright Disclaimer Under Section 107 of the Copyright Act 1976, allowance is made for “fair use” for purposes such as criticism, comment, news reporting, teaching, scholarship, and research. Fair use is a use permitted by copyright statute that might otherwise be infringing. Non-profit, educational or personal use tips the balance in favor of fair use.”

Advertisements

Count the number of discrepancies per procedure – OracleClinical (OC)

Let’s now write a quick program to count the number of discrepancies per procedure in OC/OCRDC:

Remember to comment /**/ or ***comment here*; what the program does. It is a good clinical practice to document everything so anyone can read your program and make the necessary updates, if necessary.

proc sql;
connect to oracle(path=ocpath);
create table discr as select * from connection to oracle
(Select  p.name, pd.test_order_sn detail, count(pd.test_order_sn) count, p.procedure_id procid
from discrepancy_management dm,
procedures p,
procedure_details pd
where dm.clinical_study_id=9999
and dm.procedure_id = p.procedure_id
and dm.procedure_detail_id=pd.procedure_detail_id
and p.PROCEDURE_VER_SN=pd.PROCEDURE_DETAIL_PROC_VER_SN
and dm.PROCEDURE_VER_SN=p.PROCEDURE_VER_SN
and dm.de_sub_TYPE_CODE=’MULTIVARIATE’
group by p.name, pd.test_order_sn, p.procedure_id
order by count(p.name)desc
);
/*document your code*/
proc sql;
connect to oracle(path=ocpath);
create table name as select * from connection to oracle
(select distinct p.procedure_id procid, p.name, pd.TEST_ORDER_SN detail
from  procedures p,
procedure_details pd
where p.clinical_study_id= 9999 *replace with your studyid;
and p.procedure_status_code !=’R’
and p.procedure_id=pd.procedure_id
order by procid
);
quit;

/* merge # of discrepancies with name */
proc sort data=discr;
by procid;
run;

proc sort data=name;
by procid;
run;

data discname;
merge discr (in=d) name (in=n);
by procid;
if n;
run;

proc sort data=discname ;
by descending count ;
run;

/* print out  */
proc print data=discname label;
var name numdisc percent numdcf;
label numdisc = ‘Number of discrepancies’
numdcf = ‘Number of DCFs’;
title “Number of discrepancies per Procedure”;
title2 “RA eClnica”;
run;

You could also export the report to Excel xls and have your DM / data manager review it.

Good luck and let me know if it was helpful.

To hire me for services, you may contact me via Contact Me OR Join me on LinkedIn

Professional Timeline – Clinical Programmer

Professional Timeline

Curriculum Vitae
CV

 

anayansi gamboa

Clinical Trials Terminology for SAS Programmers

Entry Level SAS Programmers

Statistical Programmer:requires him to program using the SAS language to analyze clinical data and produce reports for the FDA

Bioanalyst, Clinical Data Analyst, Statistical Programmer Analyst and SAS Programmer: same as Statistical programmer.

Biotechnology:companies which is a general term used to explain a technique of using living organisms within biological systems to develop micro-organisms for a particular purpose.

protocol:outlined all the procedures and contained detailed plans of the study.

controlled experiment: the clinical trial had patients grouped into different groups such as those in the placebo controlled group which had no active drug. This is how comparisons are made within the controlled clinical trial CFR Part 11:Code of Federal Regulations set by the FDA to regulate food, drug, biologics and device industries. The part 11 specifically deals with the creation and maintenance of electronic records.
Case Report Form or CRF:forms to collect information such as demographic and adverse events. Source Data or the information collected:which include important documents because they contain the core information required to reconstruct the essential capital of the study.
sponsor:company who is responsible for the management, financing and conduct of the entire trial. randomized: subjects that are randomly assigned to groups so that each subject has an equal chance to be assigned to the placebo control
baseline: subjects are assigned to their drug change from baseline:analyses that measure differences between baseline and current visit
placebo or sugar pill:is an inactive substance designed to look like the drug being tested. blinded:they do not know if the drug that they are taking contains the active ingredient.
open-label study:all was out in the open, the drug the subject is assigned to. Pharmacokinetics or PK:analysis of that study showed that with that dosing level, there were high levels of toxicity in the subject.
informed consent: described all the potential benefits and risks involved. TLGs: Tables, Listings and Graphs
trade name:drug name that is collected from the patient and recorded into the source data. For example: Tylenol generic name: refers to its chemical compound. For example: Acetaminophen.
WHO-DRUG: list all the drug names and how they matched to the generic drug names.This dictionary is managed by the World Health Organization MedDRA:This is short for Med (Medical), D (Dictionary), R (Regulatory), and A (Activities).
SAP: Statistical Analysis Plan ANOVA: analysis of variable
confidence interval:gives an estimated range of values being calculated from the sample of patient data that is currently in the study. null hypothesis:lack of difference between the groups in a report
pilot study:perform the same analysis upon an older. DIA: Drug Information Association
CBER: Center for Biologics Evaluation and Research (medical device) CDER: Center for Drug Evaluation and Research (drug)

Source:CDER Acronym List


Anayansi Gamboa has an extensive background in clinical data management as well as experience with different EDC systems including Oracle InForm, InForm Architect, Central Designer, CIS, Clintrial, Medidata Rave, Central Coding, OpenClinica Open Source and Oracle Clinical.