Investigating pharmaceutical product recalls due to discolouration: a Kenyan case review (January 2016–December 2024)

ABSTRACT

Product discolouration is a visually detectable quality defect that can undermine patient trust, often leading to non-adherence due to fears of contamination or chemical degradation. A retrospective study of product recalls due to discolouration, within Kenya, between January 2016 and December 2024 was conducted. Study data was extracted from the Pharmacy and Poisons Board online database, which forms a repository of all product recalls and rapid/safety alerts. All recall entries due to product discolouration were included in the study. A total of 36 products were recalled due to discolouration during the study period, with the number of recalls varying from none to 9 products per year. Tablets constituted 55.6% of the recalled products. A total of 38.9% of the products were anti-infectives. This is a notable concern, as abrupt disruption of anti-infective treatment and/or substandard products can drive drug resistance. Generic formulations comprised 97.2% of recalled products. Twenty-one industries were involved in the manufacture of recalled products, with six industries being affected by multiple recalls. Products locally manufactured in Kenya accounted for 63.9% of the recalls. A comprehensive discussion of the probable causes of the observed discolouration, guided by use of the fishbone (Ishikawa) diagram, revealed three sub-themes namely chemical reactions: microbiological and physical contamination. Product recalls due to discolouration proved to be an emerging concern, affecting up to 70 batches of a product in some cases. The in-depth discussion provides guidance on conducting investigations to identify the root cause of observed defects. The adoption of the fishbone tool alongside development of structured forms for defect reporting and quality assurance workflows is recommended.

Introduction

Pharmaceutical product recalls are a challenge with far reaching effects on the industry (damaged brand reputation, economic loss, exposure to liability), the patient (mistrust of the product, abrupt interruption of medication regimens with adverse treatment outcomes) and the healthcare system (product shortages). It is imperative that during product recalls, there is close collaboration by multiple stakeholders to ensure alternative products (alternate brand of the drug; alternate drug in the pharmacological class) are critically selected. Algabbani et al. (2023) reported an increase in drug–drug interactions in patients seen at a tertiary Saudi hospital after pantroprazole generics were recalled from the market. Physicians opted for substitution with esomeprazole or omeprazole, proton-pump inhibitors more likely to be incompatible with other drugs.

With digital platforms and social media, product recall alerts spread rapidly worldwide, leading to heightened public awareness of the prevalence and risks associated with substandard products. There has been a consistent increase in the number of pharmaceutical product recalls, affecting a wide range of pharmacological classes, over the last decade. The recent global recall of various antihypertensives of the ‘sartan’ class, due to contamination with nitrosamine a potential carcinogen, illustrates the ubiquitous nature of product recalls (Miglani et al., 2022).

The pharmaceutical industry has the impetus to predict and prevent product defects through institution of quality manufacturing systems (QMS) (ICH Guidance for industry, Q10 Pharmaceutical Quality System https://www.fda.gov/media/71553/download). QMS components as outlined by International Council for Harmonisation (ICH) Q10 include process performance and product quality monitoring system, corrective action and preventive action (CAPA) system, change management system and management review. The focus to build quality into the product (Quality by design) appreciates that all unit operations (critical manufacturing stages) need to be closely monitored. Deviation from process parameters at any stage can affect the final pharmaceutical product.

Despite requisite controls, product defects may occur thus necessitating recalls. Product recalls may be initiated by the manufacturer or the regulatory authority in response to consumer complaints about products in the market or on the detection of non-conformities in retention/stability samples. Depending on the level of seriousness of the issues leading to the recall and the risk of harm to the health of patients and the public, recalls are stratified into three classes, with class I being the most severe and likely to cause patient death (Guidelines for recall of medical products and health technologies; PPB, 2022). The extent of the recall may affect only one batch or all batches of a pharmaceutical product. The number of the affected batches is often indicative of the extent of the aberration.

Product recalls may be broadly classified into manufacturing affined and safety/efficacy affined (Nagaich & Sadhna, 2015). Among the most alarming defects to the patient is the visually appreciable changes in the formulation particularly discolouration (Khattak et al., 2018). Discolouration of medicine is likely to precipitate intentional non-adherence to treatment (Kesselheim et al., 2013), due to fear of adverse effects if patients attribute the colour change to the presence of toxic compounds. Indeed several literature studies report the discolouration of pharmaceutical products due to chemical instabilities or presence of harmful microbes (Kauffman et al., 2013; Sakiroff et al., 2022; Stark et al., 1996). There is a paucity of studies that focus solely on recalls due to discolouration, with many studies investigating all causes of product recalls. Within Sub-Saharan Africa, the assessment of product recalls in Rwanda (2019–2023) revealed a total of eight discoloured products (Bahizi et al., 2024), while a study in Zambia (2018–2021) reported a total of six discoloured products (Chabalenge et al., 2022).

The current study assessed product recalls due to discolouration in Kenya from January 2016 to December 2024. The lack of systematic national-level analysis on product recalls specifically attributed to discolouration is addressed. A critical discussion of probable causes of the discolouration and guidance to the investigative assessment is presented.

Product recalls due to discolouration: the Kenyan scenario

The retrospective study involved the collection of data on product recalls due to discolouration from the Pharmacy and Poisons Board (PPB) website (last accessed 6 January 2025). All recall entries due to product discolouration were included in the study. The constantly updated website (PPB website List of product recalls and withdrawals https://web.pharmacyboardkenya.org/recalls-and-withdrawals/) comprises a publicly accessible database of product recalls within Kenya.

For each product recall, the data extracted included the product dosage form, active pharmaceutical ingredient/s, strength, number of batches affected by the recall, manufacturing industry, country of manufacture, discolouration reported and any other concurrent defects. The data was collated into Microsoft Excel 2016 (Microsoft Corporation, Redmond, WA, USA) for further analysis. Statistical calculation of Fisher's exact test to determine the relationship between manufacturer (local vs. foreign) and nature of product recall (discolouration vs. no discolouration) was performed using Stata version 13.

A fishbone (Ishikawa diagram) (Ishikawa & Loftus, 1990; Luca, 2016), a critical quality evaluation tool, outlining probable causes of the observed product discolouration was plotted. The fishbone categories were derived from the 6Ms of manufacturing defined by Ishikawa and Loftus (1990). Emerging clusters/sub-themes drawn from the fishbone diagram, alongside relevant scenarios from the literature were reviewed.

As the study utilised publicly available, non-identifiable data from a national regulatory database, ethical approval was not required.

A total of 36 recalls, due to product discolouration, were conducted between January 2016 and December 2024 (Table 1). The product recalls were stratified according to the year of recall, product dosage form, pharmacological classification of the active pharmaceutical ingredient/s, number of batches affected by the recall, chemical classification of the active pharmaceutical ingredient/s and the product's country of manufacture.

Table 1.

Pharmaceutical products recalled by Kenyan Pharmacy and Poisons Board due to discolouration (January 2016–December 2024).

S/No	Year of recall	Pharmaceutical product	Batches affected	Country of origin	Discolouration reported	Other defects reported
1	2017	Amoxicillin and clavulanic acid powder for reconstitution 228.5 mg	One	India	Colour change	–
2	2017	Ferrous sulphate tablet	One	India	Colour change; dark spots on tablets	–
3	2017	Paracetamol tablet 500 mg	Two	India	Moulding of tablets	–
4	2017	Sulpadoxine, pyrimethamine tablet	One	Kenya A	Colour change	–
5	2017	Diclofenac gel	All	Kenya A	Product failed to comply with description	Product failed to comply with assay and related substances
6	2018	Metronidazole tablet 200 mg	One	China	Colour change	–
7	2018	Salbutamol tablet 4 mg	One	Kenya B	Inconsistent colouration; dark brown spots	–
8	2019	Chlorpheniramine syrup	One	Kenya	Colour change
9	2019	Metronidazole tablet	Two	Kenya B	Black spots on tablets	–
10	2020	Warfarin tablet 5 mg	All	Kenya C	Colour change	Hard to break
11	2020	Clofazimine soft gel capsule	Two	Switzerland	Loss of coat colour	Clumping
12	2020	Magnesium carbonate, Sodium carbonate syrup	Two	Kenya B	Colour change	–
13	2020	Chlorpheniramine, diphenhydramine, promethazine, ephedrine, ammonium chloride, sodium citrate, menthol syrup	One	Kenya D	Colour change	–
14	2020	Paracetamol tablet	Two	Kenya	Turns black; moulding	–
15	2020	Vitamin C tablet 100 mg	Three	Kenya E	Colour change
16	2020	Aminosidine tablet 250 mg	Two	Kenya A	Change in appearance	Incomplete pack
17	2021	Trimethoprim, sulphamethozaole suspension	One	Kenya B	Colour change	–
18	2022	Methyldopa tablet 250 mg	Two	Kenya B	Discolouration of tablets	–
19	2022	Povidone iodine mouthwash/gargle	Five	Kenya B	Discolouration of product	–
20	2022	Tenofovir disoproxil fumarate, lamivudine, dolutegravir tablet 300 mg/300 mg/50 mg	Seventy	Kenya A	Black spots on tablets	Discolouration of induction seal; broken tablets
21	2023	Methydopa tablet 250 mg	Three	India	Colour change; moulding,	tablet cracking
22	2023	Fluconazole tablet 200 mg	Twenty nine	Kenya A	Pink tablets displayed white discolouration	Polyvinyl chloride (PVC) and unit boxes had dark stains
23	2023	Albendazole tablet 400 mg	Six	Kenya E	Change in colour, presence of brown dots, moulding, mould like dots appear	cracking
24	2023	Ofloxacin, ornidazole tablet 200 mg/500 mg	One	India	Visible dark spots on orange colour tablets	–
25	2023	Frusemide tablet 40 mg	One	Kenya B	Black spots on tablets	–
26	2023	Amoxicillin and clavulanic acid powder for reconstitution 457mg	One	India	Colour change to brown after reconstitution	–
27	2023	Amoxicillin and potassium clavulanate dispersible tablet 312.5 mg	Twenty	India	Colour change of tablets	–
28	2024	Paracetamol solution for injection 1% w/v	One	China	Change of colour from clear colourless to yellow	White crystalline deposit at vial neck or seal
29	2024	Azithromycin suspension powder for reconstitution 200 mg/5 mL	One	India F	Colour change of suspension; black spots in suspension	–
30	2024	Elemental calcium, phosphorous, magnesium hydroxide, zinc sulphate monohydrate, vitamin D3 suspension	One	India F	Colour change of suspension; black spots in suspension	–
31	2024	Sodium chloride, potassium chloride, sodium citrate, anhydrous glucose dehydration treatment salts + dispersible tablet	Three	Kenya	Colour change	caking
32	2024	Aqueous cream 500 g	Five	Kenya	Presence of mould	–
33	2024	Paracetamol suspension 120 mg/5 mL	Four	Kenya C	Colour change from pink to brown	–
34	2024	Lidocaine and adrenaline injection	One	India	Colour change	–
35	2024	Paracetamol suspension 120 mg/5 mL	Four	Kenya C	Colour change from pink to brown	–
36	2024	Paracetamol tablets	Three	Kenya D	Dark spots on tablets	–

Alphabetic letters are used to represent product recalls from the same pharmaceutical industry; Industry B (seven products recalled); Industry A (five products recalled); Industry C (three products recalled); Industry D (two products recalled); Industry E (two products recalled); Industry F (two products recalled).

The number of products recalled per year varied from none (2016) to nine products (2024). For the subsequent calculations, the denominator was 36 (the total number of recalled discoloured products). The recalled products mainly comprised oral dosage forms namely tablets (55.6%), powders for reconstitution (11.1%), syrups (8.3%), suspensions (8.3%) and soft gelatine capsules (2.8%). The remaining products included parenteral solutions (5.6%): a mouthwash/gargle for dental hygiene, topical gel and cream each accounted for 2.8% of the total percentage. Generic products accounted for 97.2% of the recalls, and only Lamprene®, an innovator brand from Novartis Pharmaceutical Corporation was recalled.

World Health Organisation (WHO) Anatomic Therapeutic Chemical classification system (WHO ATC Anatomical Therapeutic Chemical (ATC) classification (2012) http://www.whocc.no/atc/) second-level grouping of the recalled products is presented. A total of 55.6% of the products recalled were antibacterials (amoxicillin/clavulanate; azithromycin; ofloxacin; trimethoprim/sulphamethoazole), analgesics (paracetamol; diclofenac), antiprotozoals (metronidazole; sulphadoxine/pyrimethamine) and antihypertensives (methyldopa; frusemide). The other pharmacological classes namely antivirals (tenofovir disoproxil fumarate/lamivudine/dolutegravir), antifungals (fluconazole), antithrombotics (warfarin), antimycobacterials (clofazimine), antihistamines (chlorpheniramine), antihelminthics (albendazole), vitamins, antiseptics/disinfectants, antianemics, drugs for obstructive airway disorders (salbutamol), cough and cold preparations, emollients, mineral supplements, anaesthetics, antiacids and electrolytes each accounted for 2.8% of the recalled products. Notably 38.9% of the recalled products were anti-infectives prone to the development of resistance, a significant concern to healthcare (Prestinaci et al., 2015). Critically, all the recalled antibiotic products were broad-spectrum antibiotics of critical classes such as penicillins, fluoroquinolones, sulphonamides and macrolides, product colour change leading to non-adherence and/or predicting lower potency is worrisome.

The number of batches affected by the recall varied widely among the various products: from only 1 batch to 70 batches being recalled. Notably, 41.7% of recalls affected only one batch of the respective product. The number of batches affected by a defect can give insight into the likely causes and guide the investigation pathway. A defect affecting a single batch may point to the aberration of conditions during manufacturing, whilst a defect affecting several or all batches may signify a more systemic cause.

Drug chemical classification revealed that the recalled products mainly contained amides, azoles and amines. The structural grouping termed ‘others’, thus classified due to non-standard functional group categorisation, includes vitamins, minerals, rehydration salts, povidone iodine and aqueous cream. An in-depth appreciation of the chemical structure is critical in predicting drug–excipient, drug–impurity interactions, as well as drug degradation pathways. Among the study drugs, those displaying crystal polymorphism included paracetamol, frusemide, warfarin, azithromycin, albendazole, fluconazole, amoxicillin, diclofenac and lamivudine. Polymorphic transitions may exacerbate chemical reactivity if conversion to the metastable form occurs, with the metastable form exacerbating the reaction inducing product colour change.

The recalled products were manufactured by a total of 21 pharmaceutical industries, with majority (63.9%) locally manufactured in Kenya. A significant association between local manufacture and product recalls due to discolouration (two-sided Fisher's exact test, p-value < 0.001) was observed. This is quite a considerable percentage as the Kenyan pharmaceutical industry only manufactures ∼21% of essential medicines, with market domination by imported products (Baldeh et al., 2023). This observation which suggests potential gaps in domestic quality assurance highlights the need for continual strengthening of the local manufacturing sector. The rest of the recalled products in the study were either manufactured in India (27.8%), China (5.6%) and Switzerland (2.8%). It is worth noting that the current study focused only on product recalls due to discolouration. For comparison purposes, the total number of recalled products in Kenya (January 2016 to December 2024) was 136 products. The reasons for the other product recalls ranged from mislabelling, out of specification results, mix ups, presence of impurities, leaking contents, therapeutic ineffectiveness, adverse event following administration, incomplete labelling to unpleasant taste/odour.

Six manufacturing industries (five local in Kenya) were affected by multiple product recalls, with one industry (Industry B) recalling seven different products during the study period (Table 1). The other affected industries recalled five (Industry A), three (Industry C) and two (Industry D; Industry E; Industry F) products, respectively. The higher the number of product recalls per industry, the more precarious the success of the industry and the worse the public perception.

Majority of the product recalls simply reported the defect as ‘discolouration’ or ‘colour change’ with no description of the original and final colour. This is a limitation to the study, abridging the discussion on probable causative pathways. Only 13.9% of recalls adequately elucidated the colour changes, e.g. change from clear colourless solution to yellow solution (Table 1: S/No 28: paracetamol solution). Such a descriptive report can help in cross referencing against known chemical reactions. Other reports listed ‘black spots’ and/or ‘moulding’ as the quality defect. In general, the more descriptive the report, the more insightful it is. In addition to reports on discolouration, 27.8% of reports listed other product defects such as tablet cracking, incomplete pack, dark stains on polyvinyl chloride blister and unit boxes and white crystalline deposit on vial neck/seal. These additional defects can give an indication of probable causes and areas to focus the investigation. Regional regulatory authorities can standardise defect reporting using structured forms.

Detailed approach to conducting an investigation on a defective product

Identification of any product defect warrants a thorough investigation to identify the root cause of the defect and institute CAPA. It is usually an arduous investigation, which is both time and resource intensive, exploring all probable causes.

Fishbone (Ishikawa) diagram

The fishbone diagram (Figure 1), a basic quality control tool, was used to troubleshoot likely causes of product discolouration. The brainstorming (Kumah et al., 2024) was guided under the 5Ms – Man (personnel); Machine (equipment); Mother nature (environment); Materials and Methods (process). Under each of these Ms, three most probable causes were defined. Inadequate personnel training, non-compliance with good manufacturing practices (GMP) and poor hygiene/gowning may lead to microbial contamination and/or presence of foreign matter in the final product. Equipment design, maintenance and calibration, the material used in the construction of equipment and smooth functionality of the equipment will impact on product quality. Environmental conditions such as humidity, light and temperature need to be carefully controlled to avoid hastening the chemical degradation of products. The grade, quality, purity and suitability of raw materials will play a critical role in the final product quality. Quality by design approach to manufacturing with respect to the master formula determination, process selection and validation, alongside proper institution of process parameters and controls is paramount.

Figure 1.

Fishbone diagram exploring probable causes of product colour change.

From the fishbone diagram presented, it emerges that the probable causes are clustered into three main sub-themes namely chemical reactions, microbiological contamination and physical contamination (Figure 2). These clusters are discussed elaborately below.

Figure 2.

Emerging clusters of the probable causes of the product colour change.

Discolouration of pharmaceutical products due to chemical reactions

The first index of suspicion on encountering a discoloured product is usually a chemical reaction which could be drug degradation, drug–excipient or drug–impurity interaction. To model likely reactions, the appreciation of the chemical structure of the drug/excipients/likely impurities and their functional groups is critical. The drugs constituting products recalled in the current study included amines (promethazine, adrenaline, ephedrine, diphenhramine, methydopa, frusemide, sulphamethoxazole, salbutamol, chlorpheniramine, pyrimethamine, trimethoprim); amides (paracetamol, amoxicillin, sulphadoxine, azithromycin, dolutegravir, lidocaine); azoles (fluconazole, metronidazole, ornidazole, albendazole); nucleoside agonists (lamivudine); nucleotide agonists (tenofovirdisoproxil fumarate); flouroquinolones (ofloxacin); aminoglycosides (aminosidine), riminophenazones (clofazimine), phenylacetic acid derivative (diclofenac) and coumarins (warfarin).

Drug–excipient interactions are an ever-present concern during formulation development. These interactions may cause visually appreciable changes such as discolouration or cloudiness of a product. Notably, the Maillard reaction between free amino groups and carbonyl groups of reducing sugars (e.g. lactose) results in the formation of brown–black melanoid compounds (Castello & Mattoks, 1962). Different forms of lactose are susceptible to the Maillard reaction with spray dried lactose being the most reactive. While it was not possible to get the detailed list of excipients used for manufacturing each of the products in the current study, the Maillard reaction is a plausible cause of brown/black spots reported on tablets comprising amine drugs. Knowledge of the drug chemical structure, careful selection of excipients and conducting drug–excipient compatibility studies before manufacture is recommended.

Likewise, drug–impurity interactions may also present as discolouration of a product. These impurities may range from organic compounds, heavy metals to residual solvents among others. Prachi et al. (2017) reported the instability of selected beta blockers precipitated by hydrogen peroxide impurity in the povidone excipient. It is important to ensure raw materials are purchased from prequalified suppliers who are certified free of problematic impurities. A manufacturing variation such as change of supplier or manufacturing process of the raw materials may lead to product failure due to unexpected impurities. These variations are regarded as major changes requiring process validation. Furthermore, drug–excipient compatibility studies are critical for each new batch of raw material alongside identification/purity tests of the raw material.

Drug degradation may be triggered by exposure to light, moisture, oxygen, high temperatures, unsuitable pH and/or heavy metals. The degraded drug is likely to be sub-potent and/or more toxic than the intact drug (Ellepola et al., 2020). Elucidation of the drug's degradation pathway will inform the product manufacturing process, packaging and storage conditions to preserve drug integrity. Literature reports of drug degradation associated with product colour change include frusemide tablets (Katsura et al., 2017), amoxicillin/clavulanate solution (Dawson, 2014), paracetamol and parenteral nutritional solutions (Sakiroff et al., 2022). The defective products in the current study include photolabile drugs (frusemide; amoxicillin; ascorbic acid; ofloxacin; metronidazole); moisture sensitive drugs (amoxicillin, azithromycin) and oxidation prone drugs (frusemide; amoxicillin; pyrimethamine; sulphadoxine). Drug degradation is a probable cause for some of the discolouration reports in this study, for instance the brown discolouration of amoxicillin/clavulanic acid suspension (Table 1: S/No 26) on reconstitution.

Excipient chemical reactions can also be instigated in product discolouration. The appearance of white spots on pink film-coated fluconazole tablets (Table 1: S/No 22) is suggestive of polymer precipitation associated with excessively high temperatures during coating. The concurrent presence of dark spots on the polyvinyl chloride blister and unit boxes may allude to heat damage occasioned by exposure to overheated tablets during packaging. In another study, Galí et al. (2020) observed that product temperature and drum speed during coating were variables most associated with white spots on film-coated tablets. A proportional association between product temperature or drum speed and number of defects was defined.

Critically the presence of excess moisture in the formulation of a solid dosage form, e.g. inappropriate drying of powders/granules or hygroscopic excipients, may lead to localised chemical spots due to accelerated chemical reactions such as hydrolysis or oxidation.

Discolouration of pharmaceutical products due to microbiological contamination

Microbiological contamination of a product may result in visible colour changes. In the current study, five product recalls mention ‘moulding/presence of mould’, with three reporting ‘colour change’ or turns black' alongside moulding. Though these inferences of moulding are not verified by requisite laboratory studies, they are suggestive of fungal contamination of the products. It would be worthwhile for regulatory bodies to include microbial testing data in recall documentation, when available.

Microbes (bacteria, spores, yeast) will often be introduced into a pharmaceutical product via raw materials, water, manufacturing equipment, manufacturing environment, primary packaging and/or personnel. Adherence to GMP and control of material purity is critical. The proliferation of microbes in a formulation will depend on the water activity, temperature, humidity, pH and presence of preservatives.

Microbes have been described as a kaleidoscope of colours owing to the appearance of their colonies due to the release of different pigments (Fiscarelli, 2019). The identification of bacteria or fungi by their diverse colonies is invaluable and in some cases indicates the likely contaminant of a formulation. Pseudomonas aeruginosa is associated with ‘blue-green’, Staphylococcus aureus with golden-yellow, Aspergillus flavus with ‘yellow-green’ and Aspergillus niger with ‘black’ pigments.

Kauffman et al. (2013) reported on the 2012 outbreak of Aspergillus fumigatus meningitis due to microbial contamination of methylprednisolone acetate injections compounded at the New England Compounding Pharmacy, Massachusetts. On inspection, samples of the implicated product vials were found to have green–black foreign matter or white filamentous material. Fungal contamination was later confirmed by sterility testing.

Discolouration of pharmaceutical products due to physical contamination

Physical contaminants during manufacture, e.g. metal chippings, grease from equipment; ash from charred materials; dust, hair, other foreign particles from the environment; tears, chippings from primary packing, may all appear as brown/black spots or discolouration of the products.

Discerning insights into the product recall investigations

The ability to collect information, analyse trends, investigate and identify the cause of product defects is an essential skill for any industrial pharmacist. The paper sought to provide readers with a step by step approach to investigating product defects, using the Kenyan scenario of recalls due to discolouration. Below a set of probing questions that can guide investigations are presented.

Further questions that guide investigations

1. Other than product discolouration, were any other defects observed?

The detection of other product defects alongside discolouration can be quite informative. In the current study, discolouration occurred alongside failure to comply with assay and related substances (Table 1: S/No 5 – diclofenac gel); difficulty of breaking tablets (Table 1: S/No 10 – warfarin tablets); clumping (Table 1: S/No 11 – clofazimine capsules); incomplete pack (Table 1: S/No 16 – aminosidine tablets); discolouration of induction seal, broken tablets (Table 1: S/No 20 – tenofovirdisoproxil fumarate, lamivudine, dolutegravir tablets); tablet cracking (Table 1: S/No 21 – methyldopa tablets, S/No 23 – albendazole tablets); dark stains on both polyvinyl chloride blister and unit boxes (Table 1: S/No 22 – fluconazole tablets); white crystalline deposit at vial neck or seal (Table 1: S/No 28-paracetamol solution) and caking (Table 1: S/No 31 – dehydration treatment salts).

Appearance of physical tablet defects such as tablet cracking and breaking or difficulty splitting should call for the evaluation of tabletting parameters, e.g. tablet pre-compression force, compression force, tabletting speed/dwell time, ejection force, drop height of finished tablets, facility temperature and humidity. Deviations in some of these process parameters will likely precipitate chemical reactions in the product, alongside the observed physical defects.

Discolouration of primary packaging material (Table 1: S/No 20, 22) should prompt evaluation of the packaging process parameters, e.g. sealing temperature, filling speed, facility temperature and humidity. Additionally, the suitability of the packaging material should be reassessed. Secondary packaging discolouration is more indicative of product leakages or external contamination due to unsuitable storage conditions. Evidence of deposits at the vial neck/seal of the primary packaging (Table 1: S/No 28) points to poor seal and leakage of the liquid product, with crystallisation of leaked contents. The poor seal inadvertently exposes the bulk formulation to environmental conditions, e.g. oxidation and microbial contamination.

A quality assurance report from Global fund (Global Fund QA Notice No 2023-01 V2-2023-10-04 https://www.theglobalfund.org/media/12747/psm_gfqainformationnotice2023-01-universalkenya_notice_en.pdf) summarised the for-cause investigation into defects of tenofovirdisoproxil fumarate, lamivudine, dolutegravir tablets (Table 1: S/No 20). The conclusion stated ‘The root cause for the product defects related to “Burnt Induction Seal” was due to inadequate control parameters in batch record for induction sealing machine of the bottle packing line’. The inspection was closed on 25th July 2023 with the company requalifying both the packaging equipment and process. This is the only recalled product in the current study, whose for-cause investigation is reported in the literature. Though the quality assurance report did not elucidate further on the black spots observed on the tablets, we can postulate that these spots were due to charred pieces of the burnt induction seal adhering onto the tablet surfaces. This case scenario indicates the critical nature of each unit operation through manufacturing to the final packaging stage.

• 2.How long after production did the discolouration occur?

The time within which product discolouration occurs can help postulate on the likely cause. Whilst discolouration due to physical contamination is often detectable immediately after production, changes due to chemical reactions and/or microbial contamination are generally slower and dependent on variables such as temperature, light, moisture and formulation excipients.

• 3.Did the discolouration occur in the retention samples?

The occurrence of discolouration only in market samples but not affecting the retention samples may indicate deterioration due to unsuitable storage conditions in the market. However, if even the retention samples undergo discolouration, then the problem is likely more systemic involving either manufacturing/packaging process parameters or material attributes. For retention samples, manufacturers should maintain visual stability logs and time-to-failure documentation to support investigations.

• 4.Have there been any product recalls due to discolouration within the industry previously?

Six industries were affected by multiple product recalls during the study period, with two industries reporting ‘black spots’ in more than one product (Table 1: S/No 9,25 Industry B; S/No 29,30 Industry F). Investigations to assess if the root cause of these ‘black spots’ defects are related is recommended. Industry C had two identical product recalls affecting paracetamol suspension, within a 3-month time frame, in July and October 2024 (Table 1: S/No 33,35). This raises the question as to whether adequate CAPA was carried out after the product recalls in July 2024.

Industry A recalled two different products due to discolouration associated with the packaging non-conformities namely burnt induction seal (Table 1: S/No 20) and dark stains on polyvinyl chloride blisters and unit boxes (Table 1: S/No 22). This observation calls for a critical appraisal of the packaging lines, validating the equipment, process parameters and controls.

The emergence of any pattern, with the occurrence of repeat recalls of the same product or another product with related defects, warrants a closer look to rule out similar causality.

The use of the fishbone diagram to troubleshoot likely causes, discussion of relevant studies from the literature and probing for patterns is highly informative. This fishbone tool could be formally adopted by quality assurance workflows and in product recall scenarios.

Limitations

The study design was a retrospective study using secondary data, limiting abstraction of any missing data. The recall reports did not include detailed information on the storage conditions, supply chain environments, duration on shelf nor microbial/chemical test results. Furthermore the study did not explore the relationship between product discolouration and other variables such as the manufacturing methods and/or excipients, due to the inaccessibility of this data. The study did not distinguish between contract marketing organisations and marketing authorisation holders.

The study findings may not be generalisable to other regulatory environments or low-middle income countries with different reporting standards, recall thresholds or inspection capacities.

Recommendations for further optimisation of defect reporting

A critical assessment of the discoloration defects observed and an investigative discussion on probable causes were systematically conducted by the use of the fishbone diagram. Case studies involving laboratory experiments on selected discoloured products may further test the rigour of the for-cause analysis tool. Regulatory authorities are encouraged to adopt structured defect reporting forms and standardise requisite quality tests on recalled products.

List of abbreviations

CAPA: Corrective and Preventive Action; GMP: Good Manufacturing Practices; ICH: International Council for Harmonisation; PPB: Pharmacy and Poisons Board; PVC: polyvinyl chloride; QMS: Quality Management System; WHO: World Health Organisation.

Author’s contribution

The author confirms sole responsibility for study conception and design, data collection, analysis, interpretation of results and manuscript preparation.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Availability of data and materials

All data and materials are available upon request.