Abstract
We describe the planning, selection and implementation of an advanced clinical analysis automated system into a United Kingdom National Health Service District General Hospital.
INTRODUCTION
The Princess Royal Hospital (PRH) is a district General Hospital situated in the south of England about 35 miles from London and it serves a population of about 250,000 in a semi-rural area. A regional Neurosciences Centre is attached to the Princess Royal Hospital serving a population of about 1 million.
SITUATION
The workload received in the Clinical Biochemistry department was between 300–450 routine blood samples per day with about 30% of the blood samples received needing “Immunoassay” type investigations in addition to the routine assays. All sampling was done from the original Primary Blood sample tube, currently Sarstedt Monovettes.™
Although not a particularly high workload, with the obsolete equipment in use at the time, it was very labour intensive and liable to unacceptable delays. Serious errors in resulting were not uncommon, and mechanical reliability of the equipment was poor.
Another significant problem in this area of very low unemployment is shortage of trained staff, so an effort was made to use automation to reduce the number of qualified staff needed at any one time, improving the working conditions for the staff and to enable us to spread our staff “Thinner and Longer”.
REQUIREMENTS FOR EQUIPMENT
To better understand how new equipment could improve the analytical processes within the department, we undertook a full evaluation of sample handling and result production using current equipment.
This audit of sample handling and the preliminary information from vendors enabled us to produce an “ideal” workflow schematic which was used as an aid in final equipment selection.
It became clear to us that an integrated system capable of “Routine” and “Immunoassay” techniques (ideally with some degree of pre-analytical automation) was not only appropriate, but the only viable solution for our department.
SELECTION OF PREFERRED SYSTEM
Having already made the decision that an integrated analytical system was required, we were able to consider solutions from Bayer Diagnostics™, Abbott Diagnostics™ and Roche Diagnostics™, as the only suppliers who tendered that could meet our full requirements.
The Bayer Diagnostics solution was discounted because its launch was too far in the future. The Abbott system was discounted because only the Immunoassay side was available at the time, with a dissimilar “Routine” analyser available as an interim measure.
The Roche system was selected because it had a routine chemistry system ready to operate, and the Immunoassay analyser, although not compatible with the routine analyser, was the basis for the integratable Immunoassay analyser module (using the same chemistries, racks and normal ranges) which was promised for about one year later. It was also the lowest direct tendered price. The commonality of the rack system between the Modular analyser and the Elecsys system promised the lowest sample handling, and consequently operator workload, in the first instance.
INITIAL IMPLEMENTATION
Our eventual selection was the Roche Diagnostics modular system, comprising dual Electrolyte units and dual “P” units. This would allow us to duplicate important chemistries on both P units, and also allow us considerable extra capacity.
Immunoassay capability was provided by dual Elecsys 2010 Rack analysers, as an interim arrangement, which would be superseded by the E170 modular immunoassay analyser within a period of about one year.
Routine back up capability would be supplied by a Roche Diagnostics Integra 400 series analyser and one Elecsys 2010 machine, which would be kept to provide Immunoassay back up and to enable method implementation.
This equipment was installed and commissioned in November 2000 and went “live” about five days later. Although we were not able to immediately benefit from a fully integrated system, the commonality of the Roche sample rack system between the Modular analyser and Elecsys analyser meant a simple transport between the machines – little further handling was required.
AUDIT OF WORKFLOW
The system is capable of re-running chemistries in response to pre set limits at higher or lower dilutions, and with the increased data handling now commonly available, “cascade” testing is available on the modular (e.g., analysis of Phosphate in response to a high or low Calcium results, or further automatic analysis of a liver profile in the event of a positive paracetamol (acetaminophen) result).
We have also been able to introduce “cascade” testing onto the Elecsys 2010 analysers, and we have begun this with TSH (Thyrotrophin), such that an abnormal result for TSH on Elecsys #1 will result in a request for free Thyroxine (FT4), cascading to Elecsys #2 automatically. It is also possible to set up other cascades such as Total and free Prostate Specific Antigen (PSA), and we are currently investigating this.
REPORT
Introduction of the Modular / Elecsys analyser combination has lead to improvements in our External Quality Assessment performance. We participate in the WEQAS™ QAP scheme from the University of Wales. This assesses performance by use of the Standard Deviation Index or SDI. The SDI includes components of inaccuracy and imprecision and gives a general overview of laboratory performance over time. WEQAS performance scoring of SDI is graded as <1 good; 1–2 acceptable; > 2 Poor.
On “routine” analysis, our routine SDI score has shown an improvement from an average of about 1.1 to about 0.6 following the introduction of Modular.
A similar improvement has been noted on the endocrine QAP, with our SDI decreasing from an average of about 1.0 to an average of about 0.73 following the introduction of the Elecsys.
With the introduction of the Modular / Elecsys analyser, we have been able to complete our routine workload within a standard working day with ease, which has been a benefit to staff who previously had to work into the early evening to complete the routine workload. We estimate a completion time of approximately 60–90 minutes earlier for the days “routine” workload than with our previous equipment.
Sample turnaround time is a useful indicator of efficiency, and the laboratory computer system was used to determine the turnaround times (TRT) for a batch of 40 samples for routine analytes (Sodium / Potassium/ Urea /Creatinine / Total Bilirubin / Total Protein / Albumin / Alkaline Phosphatase and Aspartate Aminotransferase). These comprise the analytes for our U/E and Liver Function Test profile. Using our previous standalone dedicated “routine” analyser, it was possible to process this batch from receipt in the laboratory until final report printing with a TRT of 92 minutes (range of 38 to 256 minutes).
The modular P units are slightly faster for the same batch of samples, taking between 21 and 207 minutes to process the batch, with a mean time of 85 minutes. This equates to a reduction of TRT of 8%.
However, the TRT taken for a similar batch of 40 U/E & Liver Function Tests but with an added request for Thyroid Stimulating Hormone, gave a mean TRT of 15 hours 47 minutes, (range of 2 hrs 47 minutes to 47 hours 50 minutes) using our old equipment. With the integrated Modular P and Elecsys system, this has reduced to a mean TRT of 2 hours 29 minutes, (range of 1 hr 34 to 4 hrs 0 minutes). This equates to a reduction in TRT of about 84%. Not only is this a substantial reduction in the mean TRT, the absolute reduction in the range of TRT is considerable.
One of our main criteria in selecting the Modular system was the ability to have fewer qualified staff on site at any one time, and this we have been able to do. The modular system has enabled us to consolidate assays from various standalone systems into one. This has enabled us to free up staff for other duties and allowed us to consider one of our design targets of working our qualified staff cadre “Longer and Thinner”.
It is now possible for one Biomedical Scientist grade to produce the sample throughput previously by produced by three. This has certainly helped in making maximal use of our qualified staff cadre. Support staff, such as lab aides, have also noticed a reduction in workload.
Cost is extremely important. We have been able to implement the Roche system with a savings of approximately 14% over the running costs of the previous free standing instrumentation. This saving in operational costs has been used to introduce newer investigations, such as Troponin T, Free PSA, and CTx crosslaps.
FUTURE DEVELOPMENTS
Whilst the selection of the Roche Modular analytical system was possible within the existing revenue budget, it was decided that acquisition of a pre analytical system would make full use of the Modular analysers' capability. The extra funding for this was found from Laboratory generated income.
Large Track systems, such as Labotix™, were studied but were found to be cost prohibitive for a laboratory of our size. Our solution was to acquire the Roche / Hitachi ™ Pre Analytical Module system (MPA), in a combination which was both useful and affordable.
We selected the Roche / Hitachi MPA standard 3, which comprises the inlet buffer module from the Modular (IBM), an Automatic centrifuge (ACU), a destopper module (DSP), an on line aliquotter module (AQN), and the output buffer module (OBM).
This is planned to commence installation in July 2001 and is scheduled to go live about four weeks after installation. The system should then reduce sample handling by staff to placing the original Primary sample onto the entry lane, taking the whole tray off the exit lane, sliding the sample racks onto the Elecsys rack carrier and later recapping samples after the second Elecsys has completed its analysis, and the sample is returned to storage.
Our final planned automation development is replacement of one of the Elecsys 2010 analysers with the Roche E170 Immunoassay analyser, which will be directly attached to the output buffer of the second P unit, thus forming a linear analytical line. It is planned that, with the introduction of the E170 unit onto the analytical line, on-line aliquots of primary samples produced at the AQN module would be directly routed to the E170, allowing parallel processing of “immunoassay” tests with “routine” tests, leading to a further decrease in turnround time and maximum throughput on the E170.
We have described our Pre-analytical and analytical systems and the benefits which we expect to arise from their introduction. We are currently investigating moving towards implementing autoverification and autoauthorisation of results, and at the present time are developing our requirements for quality automation.
The plan diagram shows the completed configuration of the Roche MPA and Modular analyser system at this site. It is designed to reduce operator fatigue and workload whilst producing high quality analysis in Clinically important turnaround times.
We believe that new generation automation is possible in the District General Hospital, and that such automation is not the preserve of the big Teaching Hospital, as has been the case previously.
After seven months of operation of the Roche Modular ISE and P units, and of the Elecsys 2010 rack analysers, we are confident of the reliability of the system, its ability to cope with our increasing workload and the further workload reduction we can expect from the introduction of the pre analytical system
Sample Rack
Pre Introduction
Post Introduction
The Plan Diagram