MAUDE Data Analytics of Smiths Medical’s CADD Infusion Pump

CADD LEGACY® PLUS PUMP

CADD LEGACY® PCA PUMP

CADD® SOLIS VIP PUMP

CADD® SOLIS PUMP

Fig.1. Different models of Smiths Medical’s CADD ambulatory infusion pumps

Study details

SPA analysts initiated a search on the MAUDE database using the Advanced Search Interface available on the USFDA site^[1]. The manufacturer’s name selected was “Smiths Medical”, and product code “FRN” (Infusion, Pump). The study period selected was ten years ranging from 1^st January 2012 to 30^th September 2021, that is, the date of receipt of report received by USFDA. During the process of retrieving information from the MAUDE database, a few challenges were encountered. For example, the database returns only the latest 500 search hits for which it provides an export option. It was observed that the records of all the adverse events reported over a period of one year are not available as a single zipped file. Therefore, for each selected year, multiple zipped files have to be downloaded from MAUDE.

For the current study covering ten years, approximately 32 zipped files were downloaded. The downloaded datasets were large, with a few files containing approximately ten million rows of data. As such large datasets cannot be processed using conventional methods, SPA leveraged Big Data Analytics to consolidate the data and convert it into a simpler form suitable for data processing. Information for the following fields was extracted from the dataset and compiled into a single spreadsheet:

1. MDR Report Key
2. Date Received
3. Brand Name
4. Manufacturer’s name
5. Patient Problem code
6. Device Problem code
7. Event Type
8. Event Description
9. Manufacturer Narrative
10. Report Number

FDA’s MAUDE database contains duplicate reports for the same faulty device or incident because of multiple reports being filed by stakeholders such as doctor, hospital, device manufacturer etc. To address this issue, SPA analysts removed the duplicate reports after compilation of data. In addition, those reports that were not relevant or those that contained improper or insufficient information were removed.

SPA analysts addressed the following questions in the current study:

1. What are the frequently occurring problems associated with the device?
2. What are the user related problems?
3. What are the critical problems that could have led to the patient’s death?
4. Are there new problems that were reported in the recent past?
5. Do opportunities exist for improving the design of the device?

Need for reclassification

MAUDE database provides classification of all the adverse events that are reported to the USFDA. Based on the data under “Event Description”, which is provided by the reporting authority, all the reports are given problem codes related to the device or patient. The device-related problem codes describe the problem faced on account of a faulty medical device during its use, whereas the patient problem code describes how the patient is affected because of the faulty device. For example, during the use of the infusion pump, the cassette should be properly attached to the pump because unattached cassette or improper cassette attachment could result in unregulated gravity infusion of medication from reservoir or a reflux of blood which could cause serious injury, resulting in death of the patient. In such cases, during data reclassification, improper cassette attachment should be considered as the device problem, and serious injury or death of the patient as the patient problem.

Fig. 2. MAUDE classification top 20 problems related to device and patient

Fig. 2 depicts the top 20 problems related to device and patient as classified by FDA’s MAUDE database.

The terminology used for describing the adverse events by MAUDE database is general in nature. For example, one of the major problems with infusion pumps is identified as “Output problem”. One cannot understand the exact nature of the problem associated with the device just by reading the classification provided by MAUDE. Also, MAUDE classifies similar device problems under different codes; for example, defective alarm and false alarm are related to alarm issues, but MAUDE has classified them as two different problems. Example screen shots from MAUDE database are presented below.

Report Number: 3012307300-2021-00492

Report Number: 3012307300-2020-09789

Also, in the screen shots presented below, under “Event Description” the device problem is a downstream occlusion. However, in MAUDE database, while one device problem has been classified under the device problem code “Device Alarm System”, the other one has been classified under “Output Problem”. Such improper classification can be seen in patient problems too, wherein the same patient effect is classified under different patient problems. Hence, there is a definite need for reclassification of adverse event reports appearing in the MAUDE database, so that one can understand the precise problem existing with the device/patient.

Report Number: 3012307300-2021-07165

Report Number: 3012307300-2018-08123

For purposes of this study, a Smiths Medical’s CADD® ambulatory infusion pump was selected by SPA analysts. The adverse event reports for Smiths Medical’s CADD® ambulatory infusion pumps were downloaded and collated, for the period 2012 to 2021.  Duplicate reports and those that were lacking in complete information were deleted. The number of adverse event reports thus retrieved was 11,010. The adverse event reports were reclassified using VantagePoint, a powerful desktop text-mining tool that helps the user analyze and visualize search results into patterns and relationships. VantagePoint also helps one rapidly navigate through a large set of results.

Insights from the study

The analysts at SPA performed the reclassification using keywords obtained from the “Event Description” text data. The number of adverse event categories after reclassification resulted in categories numbering approximately 30 for device problems, and 40 for patient problems, as compared to 280 for device problems, and 140 for patient problems in MAUDE database. Presented below are the problems identified by SPA analysts, based on reclassified data obtained using VantagePoint.

Fig. 4. Top 20 problems related to device and patient, as identified by SPA

DEVICE PROBLEMS:

Cassette attachment/damage:

From Fig. 4, it is evident that 15% of device-related adverse events were due to cassette attachment/damage that could have happened because of events such as the cassette being unattached or not properly attached to the pump, or the wrong cassette being attached, or occurrence of leakage/damage after the cassette was attached. Hence, one of the major issues associated with CADD® ambulatory infusion pumps is the user error while connecting cassette to the pump.

System failure:

The second major problem with CADD® ambulatory infusion pumps is ‘System failure’. About 11% of the reported adverse events were related to system failure that were caused by occurrences such as ‘pumping stopped’, ‘pump died’, and other malfunctions that occurred during usage.

Downstream occlusion:

There are also other adverse events associated with CADD® ambulatory infusion pumps, such as downstream occlusion which accounted for approximately 9% of the reported adverse events. When an IV becomes obstructed, the pressure within the IV syringe and connected tubing tends to increase. When the pressure reaches its setting threshold, a downstream occlusion alarm gets activated. Downstream occlusion (DSO) sensor alerts the patient and clinician if there is an interruption in fluid delivery. When there is malfunctioning of the DSO sensor, it cannot detect downstream occlusion. DSO sensor malfunction could occur because of problems such as bubbles around the seal of DSO, or cracks forming at the DSO sensor seal. The faulty DSO sensor, therefore, should be replaced before operating the device.

Volume/accuracy test:

Some CADD® ambulatory infusion pumps seem to have failed the volume/accuracy test even before they were connected to the patient. This accounted for 7% of the total adverse events. The volume test, accuracy test, and flow rate tests are conducted with multiple cassettes, and when an infusion pump fails these functional tests, it causes under or over infusion of medication. Sometimes the pump shuts off during functional tests because of faulty main processor board necessitating its replacement. Malfunction of microprocessor/circuit board also led to software timing issue, fluid ingression, processor board corruption, and faulty printed wire assembly (PWA) board.

PATIENT PROBLEMS:

No patient involvement:

Fig. 4 shows that in 90% of the adverse events that occurred during usage of CADD® ambulatory infusion pumps, the patients did not face any adverse events. Also, they were not involved in pump calibration. However, in Fig. 2 which shows problems classified by MAUDE, there are 3169 cases falling under ’No patient involvement’, while SPA analytics showed that no patient involvement is in 1956 cases only.

No adverse effects on the patient:

As in the case of device problems, MAUDE has classified identical patient problems under different headings, for example, “No consequences or impact to patient” and “No known impact or consequences to patient”, which actually mean the same thing. Below is an example of the same:

Report Number: 3012307300-2021-01193

Report Number: 3012307300-2018-03023

Report Number: 3012307300-2021-00529

There were also minor problems such as nausea, vomiting, itching, headache, pains, and shortness of breath in some patients. Patient death too was reported in 34 cases. Unfortunately, for many of these cases, it was not evident if infusion pump was the cause of their death. In some, death might have occurred due to patient’s own adjustment of the drug dosage or the practitioner’s inaccurate dosage programming. In such cases, the manufacturer’s investigations are still in progress.

User error:

Some of the adverse events reported due to user error are as follows:

• Fluid ingression into the pump by user during use.
• Failure to follow warnings, cautions, and instruction properly which resulted in serious injury to patient and damage to the pump.
• Improper attachment of cassette while connecting to pump.
• Kinks in the tube led to high pressure in the pump.
• Nurse programmed the pump incorrectly that led to death of a patient.
• Over-infusion caused by hospital staff. Pump software was compatible with Windows XP operating system only, but they did not have compatible computer, so they were unable to update the library to include the drugs.
• Pump alarm volume was set too low by the customer, and the error code displayed by the pump was “CANNOT HEAR ALARM”.
• Physical damage to the pump by customer led to microprocessor board replacement.
• Patient accidentally pushed the extra dose button on the pump.

Though only a few users error-related adverse events were reported to FDA on CADD® ambulatory infusion pumps, there could still be a research opportunity for Smiths Medical to improve the pump design, since some of the user errors can be mitigated through design enhancements. One such example is to have water resistant user interface/LCD display to function even if there is fluid ingression into the device.

Challenges:

Challenges encountered by SPA analysts during the course of this study are listed below:

1. Identifying the product code for infusion pumps:
   1. The list of product codes in MAUDE database is very long and therefore, it is difficult to identify the relevant codes.
   2. Sometimes, there are multiple product codes for the same device.

2. Downloading the entire database since 2012 and compiling the sheets:
   1. MAUDE database provides an option to download only 500 results from the search interface. So, the analyst needs to download multiple zip folders for each year and compile all the data in a single spreadsheet.
   2. In some files, the data is so large (some files have more than 15 million rows) that Excel cannot handle the data.
3. Filtering the results with respect to product code and assignee:
   1. Excel filters would not work on account of the large dataset and would often keep crashing.
   2. Duplicate reports pose a major problem while compiling/analyzing the data. MAUDE does not differentiate between adverse event reports filed by multiple stakeholders for the same device. Hence, analysts have to identify duplicate reports by considering Event date, LOT number, Catalogue number etc.
   3. Further, company names are not normalized in the adverse event reports, and so one needs to consider all possible variations of a company’s name.
4. Compiling patient problem code, device problem code, event description, manufacturer narrative etc.
   1. As the entire data is not available in a single file, there is a need to compile data related to patient problem code, device problem code, event description etc., from multiple files.
   2. Extracting definitions for patient problem code and device problem code poses a challenge.
5. Analysis and interpretation
   1. Even after all the data is compiled in a single file, the MAUDE data cannot be analyzed or interpreted meaningfully, as the same problem is categorized under different codes No standard categorization protocol is followed in MAUDE.

For reasons mentioned above, there is a need for reclassification of adverse event data in the MAUDE database. To provide clarity, SPA analysts read and reclassified the MAUDE adverse event reports.

Discussion:

Even though there are a large number of adverse events reported, 90% of the time there has been no severe complications with the patient indicating that CADD® ambulatory infusion pumps are the safest ambulatory infusion pumps available in the market. In many of the cases, manufacturer investigated the device after the receipt of the complaint from customers, and replaced the faulty components if the root cause of the issue was indeed as stated by customers.  In some instances, the manufacturer reported that there was no problem with the device, because the actual problem happened due to wrong usage by the user.

Recently, as per USFDA product recall site^[2], there was a product recall of Smiths medical’s CADD pump for faulty air-in-line detector. Smiths Medical recalled the CADD – Solis VIP ambulatory infusion pump on October 19, 2021 citing the reason that the air-in-line detector of the pump could not distinguish drug from air bubble in line and so it did not sound an alarm to notify the clinician or the user.

Smiths Medical has been continuously filing patent applications claiming enhancements to the existing product portfolio of CADD® ambulatory infusion pumps. The company has recently filed patent applications for programming method for  electronic medical record  (EMR) (WO2021243357A1), bistable display (US20210012875A1), wireless connection (US20180015218A1), push alerts (US20180322948A1), and touchless user interface (US20160228633A1). Smiths Medical’s patent portfolio for ambulatory infusion pump is robust. Some of its US patents and patent applications have been assigned or out-licensed to Tandem Diabetes^[3]. This license affords Tandem Diabetes exclusive access to technologies for future development of ambulatory infusion pumps and related software and accessories.

A medical product can become successful only if it works well and causes minimum problem during usage by the care giver or patient. By conducting analytics such as this, companies can keep track of their own products’ performance, and can also track that of their competitors. Such studies can be conducted on any type of medical device as MAUDE provides data for all medical devices such as robotic devices, cardiovascular devices, endoscopic devices, injectors etc. The findings will help identify design improvements that can be incorporated into future products.

Conclusion:

With the adoption of AI and big data analytics in all domains, it should be possible to interpret the adverse events from MAUDE database more meaningfully. Better interpretation of device problems and patient problems will help companies improve the design of medical devices resulting in overall reduction of adverse events and enhanced patient comfort.

References:

[1] https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfmaude/search.cfm

[2] https://www.accessdata.fda.gov/scripts/cdrh/cfdocs/cfRES/res.cfm?id=189372

[3]https://www.prnewswire.com/news-releases/tandem-diabetes-care-acquires-select-patents-and-patent-applications-from-smiths-medical-189359761.html