FDA 483s

Below are some recent summaries of FDA 483s from 2010 and 2011:

API maker Cambridge revises procedures, conducts training to address FDA observations

Cambridge Major Laboratories, Germantown, WI, Minneapolis District

FDA investigators Charles Cote, Sandra Hughes and Joel Hustedt inspected two facilities of Cambridge Major Laboratories, an API manufacturer, during July and August 2009, and issued each the Germantown, WI-based plant a 483.

Hot May 11 Webinar – Effective Quality Systems Compliance – FDA Expectations of Sr. Management
Featuring Acclaimed FDA Compliance Speaker John Avellanet, Cerulean Associates LLC
Save 50% -Today Only!
Register Now and Save 50% w/Offer Code 1012199

The 483 the company’s Edison Drive facility received referenced two observations. The FDA team found the company’s system suitability parameters for release testing of USP material (isradipine) did not follow the existing USP requirements for system suitability.

In its written response, Cambridge noted: “In review of the methods reliability with reduced system suitability requirements, CML has an extensive batch history to claim that the system was functioning properly at the time of analyses.”

However, the company continued: “In an effort to meet the USP requirements for system suitability, CML has amended its internal test method (TM160) to include full system suitability requirements as defined in the USP for release testing of USP material isradipine. In addition, CML has evaluated its other marketed USP-labeled APIs for similar issues relating to system suitability.”

The investigators also observed that siphon pumps for raw materials were not always compatible. The firm used a specific brand of siphon pumps to transfer solvents from drums at the time of use.

“These consist of a hand pump mechanism, a dip (suction) tube and a discharge tube,” the 483 reported. “These were noted to be stored in situ (installed into the drum). Specifications for these show construction material as PE [polyethylene.] An information sheet supplied by the firm for compatibility of containers vs. solvents show PE is not compatible with ethyl ether, one of the materials used in production of several products, including dexmedetomidine HCL.”

Cambridge explained in its reply that “although ethyl ether is used in the manufacturing of dexmedetomidine HCL” that particular API “is not manufactured at CML’s Edison Drive facility but rather at the Washington Drive facility, where compatible stainless steel pumps are utilized.”

The company also stated that the polyethylene pump used for ethyl ether at the Edison Drive plant “was immediately removed from use after it was identified by the inspector.” A compatible pump is now being used that is rated as acceptable for use with ethyl ether.

“In addition, the ethyl ether drum was immediately removed from cGMP inventory,” Cambridge stated. The firm also said it was revising its procedure for equipment and glassware design “to improve the review process for ordering of ancillary equipment (such as hand pumps) before implementation of use. The system will assess the ancillary equipment’s compatibility with different solvents.”

The Washington Drive facility received a five-time 483 from the FDA team. The report stated that stability packaging did not match bulk packaging of finished product as required in the firm’s procedure for its stability program. “Fenoldopam stability is packaged in amber bottles placed in PE bags while the bulk material is only stored in amber bottles. Midodrine stability is packaged in double heat-sealed PE bags while the bulk materials is packaged in two PE bags that are cable-tied.”

In its written response to this 483, Cambridge stated that it had “Certificate of Analyses and raw data to support that bulk fenoldopam mesylate API, although not packaged within a PE bag as demonstrated with the stability study, is stable at room temperature storage conditions for up to three years. CML has retested bulk lots of fenoldopam mesylate that have been packaged in only amber glass bottles and proven that the material is stable.”

However, the firm agreed that in the future “any bulk fenoldopam mesylate manufactured will be secondarily packaged into a polyethylene bag to mimic the stability packaging.” In addition, the company said it would “further package into a polyethylene bag” its existing inventory of the API and revise the labeling “to reflect a one-year test date from the date of manufacturing,” which will require that the API be retested before distribution.

Cambridge also noted that it had revised its stability procedure to require “the description of bulk packaging to be included in all stability protocols as a built-in check system to ensure that stability packaging is matching bulk packaging of the finished product.”

In regard to midodrine HCI, the company wrote: “CML QA performed a visual inspection of the stability samples currently stored in the stability chamber for long-term storage conditions.” The company determined from this inspection that the stability samples for the 2004 validation lots “were packaged according to the Master Production Record final API packaging instructions (i.e., twist-tied poly ethylene bags.) It was discovered that this protocol had an authoring error indicating that the samples were to be in heat-sealed bags but were not packaged accordingly.”

Any future midodrine HCI will be packaged into heat-sealed PE bags, Cambridge noted.

The FDAers observed: “Trending of stability data is not adequate for the isradipine stability study.” Specifically, they noted, data trending was not performed at the end of the accelerated studies.

Cambridge replied that it had evaluated its stability data for this API to determine if any trends were identified. “According to the Stability Protocol, the study was conducted for long-term condition only as this was an annual batch placed on stability. This study did not require accelerated conditions as they were previously conducted on the validation lots. Based on the review of long-term data, no trends were identified.”

All other studies related to isradipine were also reviewed and no trends were identified in those, as well, the firm noted.

The company told FDA that it had revised its stability procedure “to include instruction on trending and how it is reported.” Future trends analysis “will consist of graphing/ charting of the raw data” and further instructions were added to the procedure “to identify measures to be taken if a trend is identified,” including investigation, CAPA and/or generating additional reports.

FDA also faulted the company for trending only for a limited number of time points, the team noted in the 483. Again, Cambridge referenced the stability protocol that required a study only for long-term conditions for midodrine HCI. The corrective and preventive actions noted for the previous observation were applied to this observation as well.

The investigative team observed that CML’s supplier qualification was not adequate, noting that “the critical starting material for etonomine (etomidate)–(R)-(+)-alpha-Methylbenzylamine–has not been tested” as required by the firm’s subcontractor and vendor qualification procedure.

Cambridge replied that it had evaluated “all of its critical starting materials for all commercially available APIs and determined that the inadequate testing of (R)-(+)-alpha-Methylbenzylamine is an isolated incident and no systemic issue has been identified.” The company initiated a deviation report to further investigate the failure to test the raw material, and stated: “Preliminary results from the retesting of one recent retain lot has shown passing results within CML specifications for GC and chiral purity and consistency with the results reported by the vendor.”

The FDAers observed improperly labeled or unlabeled product in both cold and room-temperature storage. They noted that one lot of a product was sitting on a quarantine sticker, which was not attached. Other products and materials were found in cold storage with no quarantine or release labels affixed.

Cambridge replied: “Following the citing of the observation, CML immediately affixed all applicable labeling and ensured all labeling was consistent.” The company stated it also had conducted a retraining session with all its appropriate staff “to ensure that all materials are properly and consistently labeled and that labels are properly affixed to the containers.” The firm also had revised its internal audit schedule “to include more frequent inspections of the cGMP cold and room-temperature storage areas until there is assurance that the issue has been resolved.”

The investigation found that Cambridge had not conducted adequate investigations on discrepancies, as required by its CAPA plan. The team specifically noted the inadequacy of one investigation/disposition report. Cambridge replied: “It should be noted that this investigation was performed in 2006 utilizing a now-obsolete SOP.”

The following year, the company revised its investigative procedures “to include a more thorough process for performing investigations and determining root cause(s) for discrepancies.” The firm also stated that it was expanding its initial investigation in an attempt to identify root causes for earlier discrepancy.

A review of the company’s raw material inventory logbooks, which include inventory reconciliation, found that several of the notebooks were “inaccurate with no follow-up or discrepancy report or investigation made.”

Cambridge explained that “the raw material use logbooks are implicitly used for lot use trace-ability in the event of an investigation. These logbooks are not used for inventory reconciliation as the receiving logbook is maintained for inventory control.” The amounts recorded in these logbooks, the company said, “are approximations as exact amounts cannot always be determined based on the label information provided by the vendors.”

The firm stated that it had performed “a cursory review of the 2009 logbooks” and determined that the discrepancies were most often seen “with solvent/water drums in which multiple containers are received.”

To ensure that “the correct information is being documented on the form and a clear indication for the purpose of the use logbooks is defined,” Cambridge said it was in the process of modifying its procedure for raw material usage through the firm’s change control system. It also had held a training session “to address the proper notification of an identified discrepancy to a superior and/or QA in a timely manner in order to ensure further investigation is performed.”

AMPAC cited for contamination problems, failure to follow procedures

AMPACFine Chemicals, Rancho Cordova, CA, San Francisco District

The Feb. 9-19 FDA inspection of AMPAC Fine Chemicals, an API manufacturer specializing in high potency, cytotoxic and energetic compounds, revealed flaws in the facility’s control of potential contamination and failures to follow GMP procedures, FDA investigators Daniel Roberts and Robert Tollefsen observed.

Hot May 11 Webinar – Effective Quality Systems Compliance – FDA Expectations of Sr. Management
Featuring Acclaimed FDA Compliance Speaker John Avellanet, Cerulean Associates LLC
Save 50% -Today Only!
Register Now and Save 50% w/Offer Code 1012199

The FDAers found that AMPAC’s “Quality Unit has failed to establish adequate controls to minimize potential contamination of APIs during their manufacture.” During a walk-through of processing areas, “we identified numerous direct routes for potential contamination.” They found a charge-in port on a reactor vessel “open and uncovered when charging materials during their initial crude processing step,” the team wrote. This same port is again opened when charging materials during the subsequent purification step, as well.

In the area of this port, Tollefsen and Roberts observed piping, fixtures, mixer motors and other items “to be unclean, damaged, blistering from rust or blistering and having chipped paint.”

A reactor used during the makeup of a solution prior to its filtered transfer also included a charge-in port that the investigators observed to have a rusty cover and hold-down nuts; in addition, the cover had “chipped, blistering and flaking paint on it” and on the flat surface around the port.

The investigators reported that “brownish colored oil later determined to be off the shaft of the blender motor gearbox was observed dripping down the side of and also pooled up on the lower mounting bracket of the mixer motor. The location of this oil is at a level higher than the charge-in port” of a reactor, the FDAers noted, and therefore could potentially leak into the port and the API product.

During the manufacture of an API, the investigators explained, filters are used to collect process material, which is transferred out of a discharge port on the side of the filter into a plastic bag. “This process involves opening the uncovered plastic bag next to the discharge port prior to affixing the bag” to the port. When the material has been transferred, the bag is detached and is therefore open and uncovered directly below the reactor before it is sealed. As previously mentioned, the FDAers noted equipment above the reactors that was rusting or had blistering and peeling paint that could fall into the bag.

The EIR reported that AMPAC’s president, Aslam Malik, and its vice president for quality operations and regulatory affairs, Jeffrey Robinson, “acknowledged that the conditions we observed for the facilities and equipment … were not acceptable and would be addressed.” In its written response to the 483, AMPAC stated that although it had identified “mitigation measures” to eliminate potential sources of foreign object contamination “these had not been fully implemented … prior to the FDA inspection.”

The company said it had “taken immediate actions to eliminate the conditions observed during the inspection. Substantial cleaning of the process area was conducted to remove dirt, rust, corrosion and blistered or flaking paint from the various piping, fixtures, I-beams and mixer motors located above and around the identified open handling locations.”

The firm also identified several other steps it had taken “to prevent the future accumulation of dirt, rust or paint flakes and to address the specific findings” of the inspection. These procedural steps were being implemented throughout the facility, not only in open handling areas, the response continued.

AMPAC also opened a variance “to address the impact of the building conditions on the subject batches, which have been held in quarantine pending the actions outlined.” The company found no evidence that any contamination had occurred.

AMPAC was faulted for failing to fully investigate or implement corrective measures in response to contamination of API batches by paint chips and other debris over the previous year. The inspectors found two variance reports that the company produced during the first half of 2009 concerning possible contamination.

In the first incident, the AMPAC variance report stated: “After final packaging, operators observed foreign objects on top of the sieve screen. Two small chips were identified to be paint chips that could have fallen into the reactor during solid charging.” A Particle Contamination Cause Elimination evaluation was conducted for the process “to eliminate or reduce the chances of batch contamination by systematically reviewing and rating each step of the process” to identify opportunities for contamination to enter the batch.

The report concluded that while the batch in question “was visually inspected and no additional contamination was found, the possibility remains that the paint chips that fell into the batch could have broken into several smaller pieces, some of which could still be in the product.” For this reason, the firm’s QA unit rejected the batch of API.

A second incident reported that “a very small dark object floated to the top of the white slurry” during the first wash of an API batch. The company determined that this object likely entered the reactor through the open port during one of the solid charging operations.

FDA noted: “Both investigations are silent with respect to identification of and consideration for all other potentially affected API batches.” The reactor referenced in one incident was also used for manufacturing other APIs that could have been contaminated. The investigations also made no mention of “evaluating the various equipment” the inspectors had found with rust and flaking paint “as potentials sources of contamination of process materials.”

As of Feb. 9, the FDAers reported, “no effective controls had been implemented to minimize potential for contamination falling into” the reactor from the equipment above the charge-in port by ensuring that this equipment was properly maintained and “free of filth, rust, blistering paint or chipping paint.”

The team also observed that the internal investigations did not address evaluating:

* equipment above and around other processing locations for potential contamination opportunities;

* current cleaning procedures to determine if they “are suitable for washing out paint chips from all interior product contact surfaces of processing equipment to ensure no carryover from batch to batch or campaign to campaign;”

* current practices and procedures for visual inspection of equipment and process material to determine if those practices are suitable for detecting the presence of paint chips;

* in-process monitoring test methods, release methods or stability test methods to determine if they could detect paint chips in API product; or

* “possible chemical contamination that could be extracted from paint chips into API process materials during processing operations.” In response, AMPAC informed FDA that it had reopened these investigations and “retroactively addressed the various deficiencies raised in the observation.” The firm found that its review of other possibly affected batches found “that there were a combination of sufficient controls and methods of detection in place to prevent exposure and/or inadvertent release of product containing foreign matter.”

AMPAC found that the paint chips had come from the reactor and not from any surrounding piping or equipment; that its cleaning procedures were effective in removing paint chips; and that the interiors of equipment “afford adequate color contrast to enable detection of foreign objects through visual observation.”

The company also “identified deficiencies in procedures” for its investigation process, the response added, “which address the required content of an investigation report. Consequently, we will develop a procedure which shall detail specific investigation report requirements.”

Tollefsen and Roberts also found that AMPAC had failed to follow its established procedures for cleaning facilities and equipment used in API manufacturing. Their review of a cleaning log revealed that a processing room had not been cleaned for more than seven weeks, and an obsolete cleaning log form had been used to document cleaning in one of the production buildings.

AMPAC replied that it had opened a variance to investigate the failure to conform to cleaning procedures and found that the failure to clean the production room as required “was an isolated incident and not indicative of the overall documentation of facility cleaning throughout the plant.”

The firm also explained that although the cleaning procedure had been revised, the form was unchanged, but the blank pages in the logbook had not been updated with the correct version number.

To address these issues, the company revised its cleaning procedure to establish requirements regarding the monitoring of facilities and logbooks for cleaning activities, and that all its production, QA and Operational Excellence personnel had been retrained regarding cleaning. In addition, a stand-alone form to document cleaning was created with its own version control.

The FDA inspection revealed that AMPAC’s QA and production personnel had not followed established procedures for release of processing areas for use in manufacturing APIs. The investigators noted that the company’s Facility Readiness Audit procedure “applies to all cGMP systems used to product material” and “defines the use of a Readiness Review Form” to be signed off on by both production and QA to release rooms for use.

However, as the inspectors had noted in a previous observation, a processing room had not been cleaned in more than seven weeks, in violation of procedure, when the production manager and QA engineer signed off and approved a Readiness Review form to approve this room for use.

Further, the packaging room in one of the facility’s buildings was used for drying operations during a recent API manufacturing campaign, but no Readiness Review form was completed to approve this room for use at any time before or after this entire campaign, the FDAers reported. No variance report was created to document this deviation and all batches manufactured during the campaign were released.

The company replied that it had completed a variance report on the issue of failure to follow cleaning procedures, as noted earlier. A variance report was also completed to document the firm’s failure to use the Readiness Review form in releasing the packaging room for drying operations.

However, AMPAC stated, “activities required for release of the packaging room, while not documented on the required form, were completed prior to the use of the room. Hence, there was no adverse impact on the quality of the listed batch.”

The root cause of the variance was found to be “inadequate details regarding the process for partial release of a facility for subsequent production.”

American Hormones slapped with 13-item 483 for multiple flaws in quality control
American Hormones, Wappinger Falls, NY
New York District

American Hormones lacked appropriate procedures, testing processes, quality control and recordkeeping systems, wrote FDA investigator Demitria Argiropoulous following inspection of the compounding pharmacy.

The company had no written procedures for production and process controls, the inspector noted. No procedures described the in-process controls and tests that were to be conducted on both in-process materials and finished product from each batch for such compounded drugs as progesterone 100mg SR capsules; testosterone lipoderm 50mg/ml and 100mg/ml gel; estradiol 0.01% vaginal cream; and comp thyroid 60mg capsules.

Certain lots of these products were distributed without any finished product release testing, as well, Argiropoulous wrote in the 483 issued after the Dec. 14-29, 2006 inspection. The lots in question “were released and distributed without the active ingredient being assayed or microbiological testing having been conducted for any products.”

Control procedures had not been established to validate the performance of manufacturing processes that could cause variability in the characte-ristics of in-process materials and drug products, the 483 reported. None of the above-mentioned products had validated manufacturing processes, nor had American Hormones validated the process for producing progesterone transdermal 100mg/ml cream, testosterone 10mg troche or testosterone triturate 25mg tablets.

The company also had failed to qualify the manufacturing equipment used to produce some products. In addition, FDA said there were no equip-ment installation, operational or performance qualifi-cation protocols or reports for equipment including an encapsulation machine, two blenders and a lab oven.

An especially important deficiency at American Hormones was the company’s lack of a quality control unit, the FDAer wrote. Responsibilities and proce-dures including the authority to accept or reject in-coming com ponents, in-process materials and finished products had not been defined in writing and assigned to an employee.

American Hormones had no written stability testing program or written assessment of stability based on testing and examination of products for compatibility of ingredients. The investigator also noted that “no products have been placed on stability.

The firm never conducted accelerated and/or long-term studies on any of its products. The company arbitrarily uses an expiry that had no supporting data to justify the expiration date.

The company also accepted reports of analysis from its component suppliers in lieu of testing com-ponents for their conformity with written specifica-tions. Argiropoulous wrote that “incoming compo-nents are not tested to ensure that the testing results reported on the supplier’s certificate of analysis (COA) are accurately reporting the components’ conformity to written specifications for purity, strength and quality.”

She specifically noted that no testing had been conducted on active ingredients for formulations of testosterone, progesterone, estradiol, levothyroxine sodium USP pentahydrate and levothyroxine sodium.

American Hormones had no written proce-dures describing how to test such ingredients to ensure their purity, strength and quality, the 483 added.

The FDAer also reported that the firm had not developed its written equipment cleaning procedures
from a validated cleaning process, nor conducted vali-dation in accordance with any procedure. American Hormones did not keep records of cleaning, mainten-ance and use of certain equipment in individual equipment logs.

The company also failed to document daily calibrations of three scales used to weigh components, nor did it maintain records of maintenance and calibration of an oven used in manufacture of its drug products.

Formula worksheets used to document batch production did not include the use of certain key pieces of manufacturing equipment in the log instructions, the inspector also observed.

American Hormones was also faulted for fail-ing to establish written procedures describing the course of
action employees should follow in the case of com-plaints, nor were there provisions for review of com-plaints by appropriate personnel to determine if a complaint warranted further investigation.

Hot – 483 just released by FDA!
Post-market reporting, quality system flaws trip up Ohm Laboratories in 24-item 483
Ohm Laboratories, Gloversville, NY, Jamaica, NY District

A July 13-Aug. 12, 2009, comprehensive in-spection of Ohm Laboratories, Gloversville, NY, re-vealed numerous 483 violations in post-market reporting, quality systems, production and packaging and labeling.

An annual report did not include information about the quantity of the drug product distributed un-der the approved application, including that distri-buted to distributors, wrote FDA investigator Kevin Gonzalez.

Specifically, “your recently filed Annual Report for ANDA 78-448, Ranitidine Hel Solution USP, 15 mglmL, dated February 10, 2009 covering the review period of December 13, 2007 to December 12, 2008.”

The firm failed to include in this report the distribution data on the subject application of all lots manufactured at the company’s site and distributed from your site. The third paragraph of this Annual Product report declared that “no product has not been manufactured or distributed during the reporting period”; although according to the Packaging Line Log and Filling Room, some lots of the subject drug product were manufactured during the mentioned review period.
Next, “an NDA-Field Alert Report was not submitted within three working days of receipt of in-formation concerning a failure of one or more distri-buted batches of a drug to meet the specifications es-tablished for it in the application,” the investigator stated.

Specifically, Ohm did not keep FDA informed of a failure of a distributed drug product not meeting the specifications for it in the application. A market complaint was received on March 23, 2009 related to particles matter in Metformin Oral Solution, “and you confirmed such after receiving the complainant sample on April 1, 2009.”
The test for clarity of the solution required that the sample should be clear and free from particles or foreign matter.
“It was not until two months later that your investigation concluded that the particles may be attributed to a worn nozzle seal on the filling machine.” Corrective actions had not been taken at the time of the FDA audit, the report noted.

The report next noted that an adequate num-ber of batches of each drug product are not tested nor are records of such data maintained to determine an appropriate expiration date.

According to Gonzalez: “There is not data in place nor maintained since inception of your company and prior to April 2008 for all products manufactured that were used for stability studies including exhibits batches in support for ANDA’s as well as for your yearly stability lots. You have stability samples stored inside the stability chambers located in the warehouse where every employee may have access, as well as in the refrigerators located inside the laboratory.”
It appeared in the audit that the firm could not guarantee 100$ that samples stored in the refrigerators were ever used in exchange of the stability samples stored inside the stability chambers, as there were no records.

In addition, there are no written procedures for production and process controls designed to assure that the drug products have the identity, strength, quality, and purity they purport or are represented to possess, according to FDA records.

“There is no written procedure for the specific use of the laboratory refrigeration set [undisclosed temperature range] stability samples for Loratadine Oral Solution exhibit batches #2070701 and #2080701 as well as Radltidine HCL exhibit batch #5750601 were observed, among others,” the report stated.
These samples were observed to be stored and arranged in the same positions as those located in the stability chamber, horizontal and vertical and no in-ventory/refrigerator log was in place to show histo-ry/movement of the materials within.

Regarding quality control, the responsibilities and procedures for the QC unit were not in writing and fully followed.

For example, an OOS test result was obtained on Feb. 5, 2009. “It appears that you avoided to per-form an investigation for the assay test result. In addi-tion you failed to perform a complete assessment for the failure of the microbiology laboratory autoclave qualification that took place in February 2009.”
These incidents were not reported to the re-spective head of departments which are not stationed at the Gloversville facility, FDA stated.
Also, “the QC Director and QA Associate Di-rector which are stationed in a different State, as well as your Plant Manager and Supervisor were not noti-fied of the failure of the Microbiology Autoclave Vali-dation.”

According to the firm’s statements, they were not informed of the events, therefore, no assessment was ever performed to determine the magnitude, im-pact and significance of such failures.

Further, the company failed to have the num-ber of qualified personnel to perform and supervise the processing of each drug product.
“There is no QC supervisor physically at your firm,” the inspector wrote. “There is a director of QC, but he is stationed in New Jersey. There is no assur-ance that the entirety of data necessary for investiga-tion requiring management review is completely eva-luated and assessed to determine if appropriate inves-tigation was conducted and that appropriate corrective and preventive actions are carried out.”

Next, “employees are not given training in the particular operations they perform as part of their function and written procedures required by GMP regulations.”

For example, the plant manager as well as the production operator failed to document the actual stirring speed in an undisclosed step for a batch in December 2008. He also was involved in the compounding operation of Ranitidine HCI Solution manufactured on Jan. 29, 2008. “This last mentioned batch was rejected because the operator as well as the plant manager failed to add the entire amount of the API required during step [undisclosed] even though they signed and acknowledged that this was properly executed.

Inspection yields 10-item 483 for Baxa
Baxa Corporation, Englewood, CO
Denver District

Baxa Corporation, which manufactures Class I and II medical devices including software-controlled infusion pumps, pharmacy compounding devices and bags and automated filling systems, had a variety of GMP deficiencies, according to FDA investigator Thai Duong, who inspected the company’s facility July 12-39, 2007.

Duong noted Baxa failed to follow its own de-sign change procedures when implementing a software upgrade for its Exacta-Mix 2400, which resulted in an urgent product recall. Software version 1.07 was found to have a defect that could cause up to 50ml in “unintended volume” being added to the solution.
Each of several design changes made to the software was evaluated independently. As a result, the company did not evaluate the overall design change in its entirety and determine whether design validation was necessary.

Duong referenced several reviews that were not conducted as required by Baxa’s procedures, and added that the verification protocols for the device’s software lacked review and approval before implementation.

Baxa’s management did not dispute Duong’s findings, and “stated that the investigation for the re-call is on-going.” A CAPA was opened to address is-sues with the design change documentation and validation issues and the company stated its intention “to revise both the design change procedure and the software development process procedure.”

Review of the software changes for the Exac-ta-Mix 2400 also “revealed that the firm failed to es-tablish acceptance criteria” for some design validations, the inspector wrote in the EIR. Again, the firm’s management agreed with the observation, initiated a CAPA and promised to correct the problem.
Duong also cited Baxa for failure to identify all the actions required to “correct and prevent the recur-rence of non-conforming product and other quality problems.”

Complaints that Baxa had received about the Exacta-Mix 2400 were attributed to a “design issue that has been addressed in their current display mod-el,” the FDAer wrote. “However, the firm has yet to address the corrective action needed to correct units in the field.”

In addition, the company did not document its justification for not implementing corrective action to units in the field. Baxa’s management told the investi-gator that the company would document its justifica-tion for failure to take actions and would implement corrective actions for units in the field, as well.

The inspection revealed that “failure investigations of complaints involving possible failure of a device to meet any of its specifications are incomplete.” Baxa had attributed a number of complaints pertaining to the Exacta-Mix 2400 to “user error,” and others to “normal wear and tear,” but “had no documented evidence to support these claims,” the investigator wrote. “I stated that in order for the firm to claim user error, they need to prove the user’s wrongdoing.” In addition, in the case of claims of normal wear, “the firm needs to have the data and/or documentation to support the claims for each specific component.”
Baxa’s management representatives agreed with the observation and initiated a CAPA to address the item.

Complaint handling procedures had not been implemented to ensure that all complaints were eva-luated to determine whether the complaint should be filed as a Medical Device Report (MDR), the FDAer wrote. Duong identified some complaints that con-tained information that the firm’s devices might have caused or contributed to a death or serious injury, but that were not reported to FDA as MDR events. Other complaints were reviewed for MDR reportability but lacked the patient information required to effectively evaluate these events.

Baxa claimed that the two incidents involving serious injuries “were not reported because the firm had concluded the incidents were caused by user error.” Duong pointed out that FDA’s MDR regulation applies to events that occurred due to a variety of causes, including user error. Again, a CAPA had been initiated to address this observation and the company proposed that all MedWatch reports it receives from hospitals and pharmacies be submitted to FDA as MDRs.

Baxa’s management also proposed to revise its customer complaint system to ensure that sufficient patient information would be available for MDR eval-uation.
Duong also observed two incidents involving serious injuries that were not reported as MDRs within 30 days of the company being notified of the incidents. The inspector also observed that an MDR was not submitted within 30 days “regarding a pharmacy staff report to Baxa that they observed the Baxa MicroMacro 23 Compounder physically pumping Sodium Acetate when Sodium Chloride was expected to pump.”

Baxa agreed that its documentation was not sufficient to prove that the reports had been filed within the required timeframe and “stated that the firm plans to revise its Medical Device and Vigilance Reporting procedure, requiring the necessary docu-mentation to show that the MDR was submitted within the timeframe.”

Baxa had also failed to maintain records of a nonreportable field correction, specifically, a upgrade initiated to correct defects in software. The investigator added that “the field correction lacks a recall strategy to ensure it was conducted properly and effectively.” Again, management agreed with the observation and promised to correct it.

Duong further found that the company had no documented evidence that quality data was being covered during management reviews, as required by
Baxa’s own procedures. In addition, MDRs and corrections and removals were not being reviewed during quarterly management reviews.

The FDA investigator also observed that Baxa did not analyze all data from its quality data sources to identify existing and potential causes of non-conforming product and other quality problems.

For both issues Baxa opened a CAPA and promised to correct the deficiencies.