Avoiding Repeat Collection of Laboratory Samples at Apollo Hospitals

Gaurav Loria
Quality Coordinator, Apollo Hospitals Group
gaurav_l@apollohospitals.com

Introduction
Laboratory services are an important component of any health care organization; an effective patient assessment and treatment processes depend on the test results laboratories produce. According to the Centers for Disease Prevention and Control, more than 7 billion tests are run each year in medical laboratories, contributing to more than 70% of all medical decisions.¹  Laboratory services can also be difficult to monitor due to these services being available 24 hours a day. One of the biggest challenges laboratories face when working with this vast number of tests is the ability to accurately label, track and process laboratory specimens throughout the lab process.

Apollo Hospitals Group, which has 47 hospitals and more  than 9,000 beds in India and elsewhere, has a process in place to avoid repeat collections of laboratory samples as a safeguard testing outcomes in its laboratories across systems, and it is a key indicator that it measures and evaluates to monitor continuous compliance. 

When Apollo received complaints from patients on the number of specimen collections were being repeated due to process error, management decided that action needed to be taken to ensure that Apollo staff reduced the instances of repeat collection. Specimen collection is a safety issue both for staff and patients, but Apollo leadership considers collection an important patient-satisfaction issue, as the patient can become irritated if, for example, they are pricked for a blood sample for a second time due to a process error. Allotment of resources is also an issue; when staff must repeat a task, staff time, lab time, and physical resources are spent unnecessarily.

Description of the Specimen Collection Process
Apollo’s specimen collection process starts with preanalysis, starting with the generation of a request for a specimen sample in the ward. Subsequent steps include
• the actual testing process in the laboratory; and
• any analytical and postanalytical processes that warrant repeat testing.

When this process is performed correctly, samples are taken from the correctly identified patient, are transported and stored appropriately at the laboratory, and the correct results are reported to the correct clinician.

Apollo staff are educated to understand that it is vital to correctly identify the patient from whom the sample is taken by ensuring that the patient’s identifying details on the request label match those on the sample tube or container. Laboratory staff have the right to refuse samples that are incompletely, illegibly, or incorrectly labeled or are damaged because of improper methods of collection, storage, or transportation.

Samples excluded from these data include irreplaceable or critical samples, such as cerebrospinal fluid or a biopsy with no identifying name or number. Those samples can be accepted by the laboratory only after the requesting physician or staff responsible for the collection of the primary sample takes responsibility for identifying and accepting the sample. In such cases, the signature of the physician or staff is recorded on—or somehow traceable to—the request form.

Data Collection Approach
Every sample that reaches the laboratory is included considered in measuring results of the specimen collection process. When a repeat specimen is required, data are captured in the Repeat Sample Collection Form (see Figure 1) at each laboratory at the time of receiving the sample from patient care areas/sample collection area.

Figure 1. Repeat Sample Collection Form

Reason for error	Number of rejected specimens
Entry level
No labeling/Wrong labeling
Change in patient identification
Wrong test ordered
Incomplete entries/wrong entries in laboratory register
Phlebotomist level
Hemolysis
Clotted specimen (improper mixing)
Low sample volume
High sample volume
Collection in wrong tube
Collection from same arm with intravenous infusion
Improper transport of specimen (delay)
Tourniquet tied > one minute
Improper sample (for example, improper sputum)
Laboratory level
Reagent or instrumental failure
Abnormal values
Loss of specimen
Transcription errors (typist)
Repeat request by doctor
Other errors
Mix-up of specimens
Platelet clumps
Few spreads present

Source: Apollo Hospitals. Reprinted with permission.

Reporting Details
Reports are made monthly by the laboratory, including the following categories:
• Biochemistry
• Microbiology
• Histopathology
• Hematology
• Molecular medicine
• Transplant immunology
• Blood bank

Those reports are distributed to the following persons and groups:
• Quality Steering Committee
• Safety Committee
• Laboratory department heads
• Medical Superintendent
• Nursing Director

Results Versus an External Benchmark
Results are measured against the external benchmark rate of 0.80%, provided by Q-Tracks, collected by the United States–based College of American Pathologists. Results of data gathered between April 2008 and February 2009 are shown in Table 1.

Table 1. Monthly Specimen Repeats Versus Benchmark 

Month	Measured percentage of repeat samples	Performance vs. benchmark value (0.80%)
April 2008	0.05	.75 better than benchmark
May 2008	0.07	.73 better than benchmark
June 2008	0.04	.76 better than benchmark
July 2008	0.02	.78 better than benchmark
August 2008	0.03	.77 better than benchmark
September 2008	0.11	.69 better than benchmark
October 2008	0.03	.77 better than benchmark
November 2008	0.03	.77 better than benchmark
December 2008	0.06	.74 better than benchmark
January 2009	0.06	.74 better than benchmark
February 2009	0.05	.75 better than benchmark
Overall  average performance	0.05	.75 better than benchmark

Source: Apollo Hospitals. Reprinted with permission.

Root Causes of Rejected Specimens
The root causes of specimen rejection are shown in Figures 2 and 3.

Figure 2. Fishbone Diagram of Root Causes


Source: Apollo Hospitals. Reprinted with permission.

Figure 3. Pareto Chart of Root Causes

Source: Apollo Hospitals. Reprinted with permission.

Evaluation of Results and Steps Forward
Despite all of the results being favorable to the benchmark, Apollo Hospitals decided to continue monitoring specimen acceptability going forward. Leadership decided that enough errors in the process were still present to warrant this ongoing evaluation.

Leadership and Apollo’s Safety Committee decided to further evaluate the specimen process using W. Edwards Deming’s Plan-Do-Check-Act (PDCA) model as follows:

Plan
• The Safety Committee discussed continuing to monitor the indicator.
• All laboratory leaders were consulted and a process to begin monitoring and recording the values was designed with their input.
• The laboratory staff were educated on the importance of the process and were trained on how to record the values.

Do
• The plan was implemented and the values recorded.
• In few laboratories the data were found to be incomplete, as staff did not enter the details correctly and there were certain 
   modifications in the data collection format.
• Further data were collected and the trends were analyzed.
• According to the trend observed, the majority of the reasons for failed specimen collection were:
   - Training issues
   - Availability of an electronic system for labeling. 
   - Doctors were not satisfied with the values provided by the laboratory. 
   - Abnormal values wherein the laboratory staff repeated the test to verify the results again.
• The Safety Committee decided to conduct training sessions for nurses on usage of vaccutainers and proper collection of
   blood. Moreover, the Committee recommended sticker printers at each patient care area so that manual entering of 
   identifying information could be avoided.
• Further consultant feedback forms were devised to gather input from doctors about the services provided and then improve
   services accordingly.
• Emphasis on instructing the patients clearly on the specimen collection process and explaining to them if any special
   conditions are required.
• Repeat specimens due to abnormal values were not included in the calculations.

Check
• The data were analyzed and the values were discussed in the Quality Steering Committee. The Committee advised that
   further data should be collected.
• The values decreased further and continued to be lower than the benchmarked value of 0.80% from Q-Tracks.
• The Quality Steering Committee noted its appreciation of the nursing staff’s efforts and recommended to arrange one more
   training session to further improve on the indicator value.

Act
• The results were expected and achieved in accordance with the benchmarked value.
• As it was decided to improve further, it was recommended to go through the cycle again and plan for the improvement on
  the basis of individual reasons and their trends.

Maintaining and Expanding Success
Reduction of the laboratory sample-rejection rate has continued in 2009. Table 2 provides those data versus the benchmark.

Table 2. Recent Monthly Specimen Repeats Versus Benchmark

Month	Measured percentage of repeat samples	Performance (benchmark value 0.80%)
March 2009	0.03	Better than benchmarked value
April 2009	0.02	Better than benchmarked value
May 2009	0.04	Better than benchmarked value
June 2009	0.03	Better than benchmarked value

Source: Apollo Hospitals. Reprinted with permission

Apollo also reports increased patient satisfaction levels—scores regarding the specimen collection process have increased from 63% to 89%. After significant improvement in the process using existing technology, Apollo expects similar or increased improvement in future specimen-collection success due to the recent installation of a pneumatic tube system for specimen transport.