Corrective and Preventive Action Effectiveness in Pharmaceutical Quality Management Systems: A Systematic Review of Root-Cause Methodologies

texts, peer-reviewed articles Promotional or vendor material in the analytical core Setting Pharmaceutical manufacturing and transferable safety-critical sectors Settings with no plausible transfer to regulated manufacturing Timeframe Material available up to 2022 Material available only after 2022 Focus Mechanism or effectiveness of methodology addressed directly Passing mention without analytic treatment 2.4 Limitations of the approach Two limitations of the approach should be stated at the outset. First, no systematic screening counts or pooled quantitative outcomes are reported, because the heterogeneity of the evidence makes such figures uninformative or potentially misleading; the value of the review lies in conceptual integration rather than statistical estimation. Second, much of the strongest empirical work on root- cause effectiveness originates outside pharmaceutical manufacturing, in healthcare and high- hazard industries. The transfer of those findings to the pharmaceutical setting is reasoned by analogy, on the grounds that the failure mechanisms described are generic to investigation rather than specific to any sector, and it is identified as such wherever it occurs (Bello, Tawose, & Adama, 2022; Boakye et al., 2020). 3. Results 3.1 Conceptual and regulatory foundations of effectiveness Any assessment of CAPA effectiveness must begin from a clear definition of what the subsystem is required to achieve, and the pharmaceutical regulatory framework supplies that definition with unusual consistency across jurisdictions. The convergence is itself analytically important, because it means that effectiveness can be evaluated against a shared external standard rather than against organisation-specific preferences. The United States framework places the investigative obligation at the centre of good manufacturing practice, requiring that any unexplained discrepancy or failure of a batch to meet specifications be thoroughly investigated, which establishes investigation, and by extension root-cause analysis, as a legal duty rather than a discretionary good practice (U.S. Food and Drug Administration [FDA], 2020a). The FDA quality-systems guidance reinforces this by describing CAPA as the element through which an organisation learns from problems and prevents their recurrence, and by linking the rigour of CAPA to the risk posed by the underlying problem (FDA, 2006). The ICH Q10 model generalises this expectation into a global standard, identifying CAPA as one of the enabling elements of the pharmaceutical quality system and stating that a structured approach to the investigation process should be used with the objective of determining root cause. Critically, Q10 introduces the principle of proportionality: the level of effort, formality, and documentation of the investigation should be commensurate with the level of risk, a principle elaborated in the companion guideline on quality risk management (ICH, 2005, 2008). The European framework expresses the same logic, requiring that the root causes of quality defects be identified, that appropriate corrective and preventive actions be taken, and that the effectiveness of those actions be monitored and assessed (European Commission, 2013). The explicit requirement to assess effectiveness, rather than merely to implement actions, is significant, because it makes verification of effectiveness a defined regulatory expectation rather than an optional refinement. Comparable requirements appear in the general quality-management and medical-device standards, and in the World Health Organization good manufacturing practices (ISO, 2015, 2016; World Health Organization, 2011), and effective operation of any such system in a highly regulated environment depends on disciplined coordination and control of the underlying processes (Eyetsemitan et al., 2020), supported by risk-based audit and assessment that surface quality signals before they escalate (Akomolafe and Agu, 2018) (Dada & Isiekwu, 2021a; Dada & Isiekwu, 2021b). Within this framework, CAPA effectiveness has a meaning that is narrower than general notions of quality improvement. An action is effective when it eliminates or adequately controls the cause it was designed to address, such that the targeted nonconformity does not recur within the relevant population and timeframe, and when this elimination can be demonstrated through objective evidence rather than assumed. Three components follow and structure the remainder of the review. Causal accuracy requires that the action target the actual root cause rather than a symptom or a contributing factor, and is the specific contribution of root- cause methodology (Atima & Sanni, 2022; Badmus et al., 2020). Action adequacy requires that the chosen action be capable, in principle, of controlling the identified cause, which is a question of design; because many corrective actions take the form of procedural controls, the design quality of standard operating procedures bears directly on adequacy, and procedures treated as strategic control assets rather than static documents are a primary vehicle through which corrective and preventive actions are sustained (Eyetsemitan et al., 2022). Verification of effectiveness requires a defined check, conducted after implementation and after a meaningful interval, that confirms recurrence has in fact been prevented (European Commission, 2013; ISPE, 2020). A failure in any one of these components defeats effectiveness regardless of performance in the others: a perfectly verified action against the wrong cause is ineffective, and an accurately targeted action that is never verified cannot be shown to be effective and, in a regulated setting, is therefore incomplete (Dagodzo, 2021a). A related source of confusion, and of ineffective CAPA, is the conflation of different categories of cause. Investigations commonly distinguish the direct cause, the immediate event that produced the nonconformity, from contributing causes, the conditions that made it more likely, and from the root cause, the underlying systemic condition whose correction would prevent recurrence. Corrective action aimed at a direct or a contributing cause may reduce the immediate effect without removing the generative condition, which is why a clear and shared causal vocabulary is itself a precondition of accurate analysis. The regulatory expectation that investigations reach root cause, rather than stopping at the most visible antecedent, is in effect a requirement to push the analysis through these layers to the level at which durable prevention becomes possible (Dagodzo, 2021b). 3.2 A taxonomy of root-cause methodologies Root-cause analysis is best understood not as a single procedure but as a set of methods that differ in the logical structure they impose on an investigation. Following the comparative tradition in the quality literature, the methods can be distinguished by their reasoning structure, which determines the kinds of causal relationships they are able to represent (Doggett, 2005; Andersen and Fagerhaug, 2006). Four broad structures recur. Linear methods trace a single causal chain backward from an event. Categorical methods organise candidate causes into predefined groups. Deductive methods decompose an event through formal logic. Prospective methods reason forward from potential failures that have not yet occurred. A fifth grouping comprises integrative frameworks and systems-oriented methods that combine these structures within a disciplined process (Rooney and Vanden Heuvel, 2004). The subsections that follow characterise the principal methods within each grouping. 3.2.1 Linear and iterative methods: the 5 Whys The 5 Whys, drawn from the Toyota Production System, is the most accessible root-cause method. The investigator states the problem and asks why it occurred, then asks why of each successive answer until a cause is reached that, if corrected, would prevent recurrence (Ohno, 1988; Liker, 2004). Its strengths are real: it requires no specialised training, it can be applied rapidly and even individually, and its iterative discipline counteracts the common tendency to stop at the first plausible explanation. For simple problems with a single dominant causal chain, it is often sufficient and is the proportionate choice. Its weaknesses are equally well documented and are intrinsic to its linear structure. Because it follows one thread, it tends to converge prematurely on a single root cause and to miss the multiple, concurrent, and interacting causes that characterise complex events. The path the questioning takes is sensitive to the assumptions of the investigator, so that two competent analysts may reach different conclusions from the same incident. Card argues on these grounds that the method, while valuable for teaching causal reasoning, is poorly suited to the investigation of serious or complex events, where its tendency to oversimplify can produce confidently wrong conclusions (Card, 2017). In a pharmaceutical context, a 5 Whys analysis that terminates at operator error, rather than at the procedural, design, or system condition that made the error likely, is a recognised pattern of failed CAPA, and one that the regulatory emphasis on systemic causation is intended to prevent (Dagodzo & Ahiaeke Patrick, 2022; Dogbatsey & Ebhojie, 2019). 3.2.2 Categorical methods: the Ishikawa diagram The Ishikawa or fishbone diagram organises potential causes into standardised categories, conventionally summarised for manufacturing as manpower, machine, material, method, measurement, and environment (Ishikawa, 1990). The problem is placed at the head of the diagram and candidate causes are mapped along branches representing each category. Its principal advantage over linear methods is breadth: by prompting investigators to consider every category systematically, it counters the premature narrowing to which the 5 Whys is prone, and it functions well as a structured group activity that surfaces a wide field of candidate causes and makes the team's collective knowledge explicit (Dosunmu, 2022; Ekeocha et al., 2021). Its limitations are the mirror image of its strengths. The diagram is a tool for organising hypotheses, not for testing them; it displays candidate causes without weighting their contribution or establishing which are operative (Andersen and Fagerhaug, 2006). It can also encourage a false sense of completeness, since a fully populated diagram may still omit the true cause if the investigating team lacks the relevant knowledge or fails to challenge its own assumptions. In practice, the fishbone diagram is most powerful when used to generate the field of possibilities that a subsequent, more analytical method then narrows and verifies, rather than as a standalone determinant of root cause (Eyetsemitan et al., 2021; Eze & Anene, 2022a). 3.2.3 Deductive logic methods: fault tree and event tree analysis Fault tree analysis decomposes a defined top event through Boolean logic, using AND and OR gates to represent the combinations of lower-level failures that can produce it (Vesely et al., 1981). Developed for high-consequence engineered systems, it is the most rigorous of the common methods and is uniquely able to represent the way multiple conditions combine to cause an event, rather than reducing causation to a single chain. Where the underlying failure probabilities are known, it also supports quantitative analysis, allowing the relative contribution of different failure paths to be estimated and the most significant contributors to be prioritised. Event tree analysis provides the complementary forward view, tracing the possible consequences of an initiating event through success and failure of intervening barriers (Eze & Anene, 2022b; Filani et al., 2022). These capabilities come at the cost of effort and expertise. Fault tree analysis requires an understanding of the system architecture and of Boolean logic, is time-consuming to construct and to validate, and is therefore disproportionate for routine, low-risk events. Its appropriate domain is the complex, high-risk, or safety-critical event where the combinatorial structure of causation must be made explicit and where the proportionality principle of ICH Q10 justifies the additional investment (ICH, 2008; Vesely et al., 1981). In pharmaceutical practice, deductive methods are most often reserved for serious events affecting sterile assurance, cross-contamination control, or other failure modes with a direct line to patient harm (Ike et al., 2021). 3.2.4 Prospective methods: failure mode and effects analysis Failure mode and effects analysis differs from the preceding methods in its orientation. Rather than reasoning backward from an event that has occurred, it reasons forward from potential failure modes that have not yet occurred, scoring each on severity, occurrence, and detectability to derive a priority for mitigation (Stamatis, 2003). This prospective character makes FMEA the natural instrument of preventive action, the forward-looking half of CAPA, and aligns it directly with the risk-based orientation of the regulatory framework and with the tools of quality risk management (ICH, 2005). It is also widely used in process and design development, where it supports the building of quality into the process rather than the correction of failures after the fact (Isiekwu & Oluwo, 2021; Ladapo, Dosunmu, & Jooda, 2022). Used reactively, the outputs of an FMEA can inform an investigation by directing attention to the failure modes a process was already known to be susceptible to, and by providing a documented baseline against which an actual failure can be compared. Its limitations are that it depends heavily on the foresight and knowledge of the team, that its numerical scoring can convey a false precision and can be manipulated to justify predetermined conclusions, and that it analyses failure modes largely in isolation rather than capturing their interaction. As a structured way to convert the lessons of past failures into anticipatory controls, however, it is the method most closely associated with the shift from reactive correction to genuine prevention (Stamatis, 2003) (Ladapo et al., 2019; Ladapo, Jooda, & Dosunmu, 2022). 3.2.5 Structured improvement cycles: DMAIC and Lean Six Sigma Structured improvement cycles embed root-cause analysis and corrective action within a disciplined, data-driven sequence rather than treating them as isolated steps. The Lean Six Sigma define, measure, analyze, improve, and control cycle is the most prominent example: it defines the problem and its impact, measures current performance, analyses the data to identify root causes, improves the process by addressing those causes, and controls the improved process to sustain the gain (Pyzdek and Keller, 2014; George, 2002). The analyse phase is itself a root- cause engine and typically draws on the categorical, statistical, and graphical tools already described, while the control phase institutionalizes the verification and monitoring that the regulatory framework requires, the cycle descending from the plan-do-check-act logic of continual improvement (Deming, 1986). Systematic reviews of Lean Six Sigma indicate that the approach improves quality and reduces variation across diverse settings, and that the cycle can be scaled or adapted to the resources and risk of the organisation rather than applied uniformly, a finding of particular relevance to smaller enterprises and to the proportionality principle (Ambali et al., 2021; Antony, 2006) (Liadi, 2022a; Liadi). The DMAIC structure is significant for CAPA effectiveness in two respects. First, by coupling analysis to measurement and control within a single cycle, it directly addresses the gap, discussed below, between conducting an analysis and verifying that the resulting action worked. Second, applied studies in this tradition demonstrate corrective action operating at the level of the process rather than the individual event, for example through structured redesign that reduces process cycle time while holding quality constant (Oyeleye et al., 2022). The principal caution is that the rigour of DMAIC depends on the availability and integrity of process data, and that the method can become a procedural ritual if its analytic phase is conducted superficially (Liadi, 2022c; Lilian et al., 2020). 3.2.6 Team-based and rational frameworks: 8D, Kepner-Tregoe, and Pareto analysis Several further frameworks combine the preceding tools within a disciplined problem-solving sequence. The eight disciplines approach embeds cause analysis within a team-based structure that spans problem definition, interim containment, root-cause identification, corrective action, verification, and prevention of recurrence, and is widely used where supplier quality and customer response are at issue (Okes, 2019). The rational decision-making tradition associated with Kepner and Tregoe contributes the is and is-not comparison, a discipline for specifying precisely what the problem is and, equally informatively, what it is not, in terms of identity, location, timing, and magnitude, thereby bounding the search for causes and guarding against unfocused speculation (Kepner and Tregoe, 1981). Pareto analysis contributes a complementary prioritization function, using the empirical concentration of effects among a few causes to direct investigative and corrective effort toward the vital few rather than the trivial many (Juran and Godfrey, 1999). These frameworks matter for the present argument because they situate tool choice within a process, reinforcing the finding, developed below, that effectiveness is a property of the investigative system rather than of any single instrument (Mbonu, Aliliele, & Iwuanyanwu, 2020; Mbonu et al., 2021a). 3.2.7 Systems and human-factors methods A final grouping reflects the influence of human-factors and systems thinking on investigative practice. Barrier or control analysis examines the defences that should have prevented an event and asks why each failed, directing attention to the layers of protection in a system rather than to a single point of failure. Change analysis compares a situation in which a problem occurred with a similar situation in which it did not, isolating the change responsible. Causal factor charting and proprietary structured methods such as Apollo root-cause analysis build an explicit map of causes and effects, requiring evidence for each causal link and thereby disciplining the reasoning by requiring evidence for each causal link (Gano, 2008; Latino, Latino, and Latino, 2019). These methods are distinguished by their insistence that serious events arise from the interaction of multiple conditions and latent weaknesses rather than from isolated errors, a position grounded in the broader human-error and safety literature (Reason, 1990; Leveson, 2011). Their cost is greater complexity and a requirement for skilled facilitation, which again locates their appropriate use in the investigation of serious or recurrent events (Mbonu, Aliliele, & Iwuanyanwu, 2022). Table 2 consolidates the taxonomy on the attributes that bear most directly on method selection. The comparison is qualitative and is intended to support proportionate matching of method to problem rather than to rank the methods absolutely; as the evidence reviewed below makes clear, no method is effective in isolation from the conditions of its use. Table 2. Comparative summary of common root-cause methodologies by reasoning structure, temporal orientation, domain of best fit, and principal limitation. Method Reasoning structure Orientation Best suited to Principal limitation 5 Whys Linear, single chain Retrospective Simple, single- cause problems; rapid response Premature closure on one cause; misses interacting causes Ishikawa (fishbone) Categorical Retrospective Multi-factor problems needing broad hypothesis generation Organises but does not test or weight causes Fault tree analysis Deductive, Boolean Retrospective Complex, high- risk, safety- critical events Time and expertise intensive; disproportionate for routine events FMEA Prospective, scored Prospective Preventive action; anticipating failure modes Depends on team foresight; treats modes in isolation DMAIC (Lean Six Sigma) Cyclic, data-driven Both Process improvement Depends on data integrity; can become ritualistic Method Reasoning structure Orientation Best suited to Principal limitation with measurable performance data Eight disciplines (8D) Team-based sequence Retrospective Supplier and customer-facing quality problems Administratively heavy for minor events Pareto analysis Frequency- based Retrospective Prioritising among many recurring causes Directs effort but does not explain causation Barrier / change / causal- factor methods Systems, evidence- mapped Retrospective Serious, multi- factorial events with latent weaknesses Complex; requires skilled facilitation 3.3 Determinants of effectiveness The taxonomy establishes that the methods differ in capability. The more consequential question is whether, and under what conditions, the use of these methods actually produces effective corrective and preventive action. Here the literature is more cautionary than celebratory, and the dominant message is that root-cause methodology is necessary but far from sufficient for effectiveness. The most influential critical appraisals come from healthcare and high-hazard industries, where the practice has been studied more directly than in pharmaceutical manufacturing, and their findings transfer because the failure mechanisms they describe are generic to investigation. Peerally and colleagues identify a series of structural problems that limit the value of root-cause analysis as commonly practised, including the political and social pressures that shape investigations, the tendency to focus on the proximal actions of individuals rather than on system conditions, the weakness of many resulting action plans, and the absence of mechanisms to confirm that recommended actions are implemented and effective (Peerally et al., 2017). Wu and colleagues raise a complementary concern, observing that the effort invested in conducting analyses is frequently not matched by the effectiveness or implementation of the actions that follow, so that the analytical exercise can become an end in itself (Wu et al., 2008). Four determinants recur across this literature. 3.3.1 Matching method to problem The first determinant is the fit between method and problem. The comparative and tool-selection literature is consistent that no single method is universally superior and that the appropriate method depends on the structure, complexity, and risk of the problem under investigation (Doggett, 2005; Andersen and Fagerhaug, 2006). An empirical comparison of three analysis tools found that they locate root causes with differing accuracy and quality, confirming that the choice of tool affects the outcome of the analysis (Doggett, 2004). A simple operational anomaly may be resolved efficiently by a 5 Whys analysis, whereas a complex event with interacting causes requires the breadth of a categorical method, the rigour of a deductive one, or a combination. The misapplication of a method to a problem whose structure it cannot represent is itself a cause of ineffective CAPA, because it produces an analysis that is internally coherent but causally incomplete. This finding aligns directly with the proportionality principle embedded in the regulatory framework: ICH Q10 expects the effort and formality of investigation to be commensurate with risk, which is in effect a regulatory instruction to match method to problem (ICH, 2008). Doggett formalises the same logic by proposing that tool selection itself be treated as a deliberate, criterion-based decision rather than a matter of habit or local convention (Doggett, 2005). The practical implication is that an organisation should possess a repertoire of methods and an explicit rationale for choosing among them, rather than defaulting to a single tool for every event (Mbonu et al., 2020b; Mbonu). 3.3.2 Triangulation for complex events The second determinant is the use of more than one method for complex events. Because the methods differ in the causal structures they can represent, their weaknesses are partly complementary. A fishbone diagram can generate a broad field of candidate causes that a fault tree or a disciplined 5 Whys then narrows and tests; an FMEA can convert the verified causes of a past event into anticipatory controls; and a DMAIC cycle can wrap measurement and control around the whole (Andersen and Fagerhaug, 2006; Stamatis, 2003; Pyzdek and Keller, 2014). The structured problem-solving frameworks formalise this combination by embedding several tools within a single investigative sequence (Okes, 2019). The literature on the failure of single-method analyses, particularly the critique of the 5 Whys, can be read as indirect support for triangulation, because many of the documented failures are failures of a single linear analysis where a multi- method approach would have exposed the missing causes (Card, 2017; Peerally et al., 2017). Triangulation is not a counsel of maximal effort; it is the proportionate response to problems whose causal structure exceeds the representational capacity of any single tool. Triangulation nonetheless carries a cost, and proportionality applies to it as it does to method selection. For routine, low-risk events a single appropriate method remains the correct choice, and the indiscriminate use of multiple methods would waste investigative resource and dilute focus. The value of triangulation rises with the complexity and consequence of the event, which is why it is most defensible for serious or recurrent problems (Mbonu, Iwuanyanwu, & Uzoka, 2019). Independent review, in which a second competent analyst examines the reasoning, provides a complementary safeguard by exposing assumptions that a single investigator may not see, and is a low-cost form of triangulation applicable even where multiple formal methods are not warranted (Michael, 2019a). 3.3.3 Human and organisational climate The third and arguably strongest determinant is the set of human and organisational conditions within which analysis is conducted. The human-factors tradition, together with frameworks developed specifically for human error causation in high-risk activities, demonstrates that the framing of causation is not neutral. The widespread tendency to attribute failure to the errors of individuals at the sharp end, rather than to the latent system conditions that shape behaviour, leads investigations toward corrective actions, such as retraining or disciplinary measures, that have weak preventive value because they leave the generative system unchanged (Obriki and Arumosoye, 2020; Reason, 1990, 2000). Hindsight bias compounds this, making the path to an adverse outcome appear more foreseeable in retrospect than it was in prospect and thereby distorting the assignment of cause (Dekker, 2014). Systems-oriented accounts of safety argue further that focusing on component failure misses the way accidents emerge from the interactions of a complex system, and that effective prevention requires reasoning about the system as a whole (Leveson, 2011; Hollnagel, 2014; Michael, 2019b; Michael, 2021a). Organisational conditions determine whether even a sound analysis yields a sound action. A blame-oriented climate suppresses the candid disclosure on which accurate analysis depends, so that investigators are told a defensible story rather than the true one; time pressure curtails investigations before contributing causes are identified; and the absence of investigator competence or skilled facilitation degrades the application of any method (Peerally et al., 2017; Carroll, 1998). The way an organisation learns from incidents is itself an organisational capability that varies widely and that can be deliberately developed (Carroll, 1998). These moderators help explain why two organisations using the same nominal methodology can achieve very different rates of recurrence, and they direct attention away from the question of which tool is best and toward the question of whether the conditions for effective investigation are present at all (Michael, 2021b; Michael, 2021c). 3.3.4 Verification of effectiveness The fourth determinant is the verification of effectiveness, the defined check that confirms, after implementation and after a meaningful interval, that the targeted nonconformity has not recurred. The regulatory framework requires this assessment explicitly, and professional guidance treats it as a defining feature of a mature CAPA system (European Commission, 2013; ISPE, 2020). Yet the critical literature repeatedly identifies weak or absent verification as a principal reason that analyses fail to translate into improvement, because actions are closed on the assumption of success rather than on its demonstration (Peerally et al., 2017; Wu et al., 2008). Verification performs two functions. It confirms, against objective evidence such as trended recurrence data, that the action achieved its purpose, and it closes the learning loop by exposing analyses that were causally inaccurate, so that an ineffective action prompts re-investigation rather than the silent persistence of the underlying condition. An effectiveness check is therefore not a bureaucratic appendage to the CAPA process but the mechanism that makes the entire process self-correcting; its omission converts CAPA from a learning system into a record-keeping one (Michael, 2021d; Michael, 2021e). 3.4 Measuring and verifying effectiveness Because verification is so central, the way effectiveness is measured deserves separate consideration. The most direct outcome measure is recurrence: whether the same or a closely related nonconformity reappears within a defined population and period after the action is implemented. Recurrence is a lagging indicator, informative but available only after time has passed, and its interpretation requires a baseline rate and an adequate observation window. Leading indicators, by contrast, provide earlier signals of whether an action is likely to be effective, for example the degree to which a new control is actually used as intended, the results of process capability studies after a change, or the closure of the specific conditions identified in the analysis (Obogo & Arumosoye, 2020c). A balanced effectiveness check draws on both, using leading indicators to detect early drift and lagging indicators to confirm durable prevention (Michael, 2022a). The design of the effectiveness check should follow from the nature of the action. A change to a procedure may be verified by audit of adherence and by trending of the relevant deviation type; a process change may be verified by statistical evaluation of capability before and after; an equipment modification may be verified by qualification and by monitoring of the failure mode it was intended to eliminate. Trending and signal detection at the level of the quality system, rather than the individual record, allow weak or recurring signals to be identified that would be invisible in a single investigation, which is why periodic management review of CAPA data is itself a control on effectiveness (ICH, 2008; ISPE, 2020). Two design choices govern the credibility of an effectiveness check: its timing and its evidentiary basis. The interval before verification must be long enough for the targeted nonconformity to have had a realistic opportunity to recur, since a check conducted too soon confirms only that the action was implemented, not that it works. The evidence relied upon must be objective and specific to the cause addressed, rather than a general assurance that no problems have been observed. Where recurrence is rare, the absence of a single repeat event over a short window is weak evidence of effectiveness, and leading indicators or process-capability data carry more weight. Designing the check at the time the action is planned, rather than retrofitting it at closure, is the practical safeguard that keeps verification meaningful (Obogo & Arumosoye, 2020b). A persistent difficulty in measuring effectiveness is attribution. Because pharmaceutical processes operate within a system of overlapping controls, a reduction in recurrence after a CAPA may reflect the action itself, concurrent changes elsewhere in the system, or simple regression to the mean, and disentangling these is rarely straightforward. The problem is most acute for rare events, where the statistical power to detect a genuine change is low, a difficulty documented for rare events in the chemical process industry (Kumari et al., 2020), and for actions implemented alongside other improvements. Robust verification therefore favours evidence that links the observed change specifically to the cause addressed, such as a measured improvement in the parameter the action was designed to control, rather than a general decline in the broad category of deviation. Where attribution cannot be established with confidence, the honest conclusion is that effectiveness remains unconfirmed and continued monitoring is warranted, rather than that the action has succeeded. Table 3 summarises the contrast between leading and lagging effectiveness indicators and their typical uses. Table 3. Leading, lagging, and system-level indicators for verifying CAPA effectiveness Indicator type Examples Use and limitation Leading Adherence to a new control; post- change capability studies; closure of identified conditions Early signal of likely effectiveness; does not by itself confirm durable prevention Lagging Recurrence rate of the targeted nonconformity over a defined window Direct confirmation of prevention; available only after time has passed System-level Trending of deviation categories; periodic management review of CAPA data Detects weak or recurring signals invisible in single investigations 3.5 Convergent evidence from other safety-critical sectors Because the pharmaceutical literature on investigative effectiveness is comparatively thin, the experience of other safety-critical sectors provides important convergent evidence. In healthcare, formal root-cause analysis became a mandated response to serious adverse events, yet sustained study found that the analyses often produced weak, person-focused actions whose implementation and effectiveness were rarely confirmed, prompting calls to strengthen the action and verification stages rather than the analytical step alone (Peerally et al., 2017; Wu et al., 2008; Vincent et al., 1998). In aviation and the nuclear and process industries, where deductive and systems methods are long established, the lesson has been that reliable prevention depends on treating events as emergent properties of complex systems and on maintaining a reporting culture in which problems are disclosed without fear of blame (Reason, 1990; Leveson, 2011; Hollnagel, 2014). In manufacturing more broadly, the Lean Six Sigma tradition demonstrates that coupling analysis to measurement and control within a disciplined cycle improves the durability of corrective action, an effect visible in applied process redesign that reduces cycle time while preserving quality (Pyzdek and Keller, 2014; Ambali et al., 2021; Oyeleye et al., 2022). The convergence across these sectors is striking. Despite very different technologies and methods, the determinants of effective prevention are consistent: match the method to the complexity of the problem, look beyond individual error to system conditions, sustain a climate that permits honest analysis, and verify that actions work. This consistency strengthens the inference that the determinants identified in the pharmaceutical literature are not artefacts of a particular setting but general features of how organisations learn from failure (Obogo & Ozobu, 2019b). The process and chemical industries contribute a further, complementary insight: the view of safety as a series of layered defences, each imperfect, such that an event occurs only when weaknesses in successive layers align. On this model, effective corrective and preventive action strengthens the defences as a system rather than targeting the single layer that happened to fail in a given event, and the investigation asks not only what failed but why the other defences did not catch it (Reason, 1990, 1997; Leveson, 2011). For pharmaceutical quality systems, in which product quality is protected by overlapping controls across design, process, and testing, this layered perspective reinforces the case for analysing serious events with methods capable of representing multiple interacting causes rather than a single point of failure. Table 4 summarises the cross-sector evidence. Table 4. Convergent lessons on investigative effectiveness from adjacent safety-critical sectors. Sector Characteristic practice Principal lesson for CAPA Healthcare Mandated root-cause analysis of serious adverse events Strengthen the action and verification stages, not the analytical step alone Aviation and nuclear Systems and deductive methods; strong reporting culture Treat events as emergent system properties; protect blame-free disclosure Process manufacturing Lean Six Sigma and structured improvement cycles Couple analysis to measurement and control to sustain corrective action 3.6 Integration with quality risk management The determinants identified above connect CAPA to the wider discipline of quality risk management, which the regulatory framework treats as a companion to the quality system (ICH, 2005, 2008). Risk management supplies both the rationale for proportionality and a set of tools, notably FMEA and risk ranking, that operate at the interface between prospective prevention and retrospective correction. A risk-based CAPA system uses the severity and likelihood of the underlying hazard to determine the depth of investigation, the rigour of method selection, and the priority and timeliness of action, so that the gravest problems receive the most thorough analysis and the most robust verification, while minor anomalies are resolved proportionately; risk-based audit and assessment feed this prioritisation by surfacing the signals that warrant deeper investigation (Akomolafe and Agu, 2018). This integration also disciplines the preventive half of CAPA. The output of an investigation into one event can be generalised, through risk assessment, to identify other processes or products exposed to the same mechanism, converting a single corrective action into a broader preventive one. In this way quality risk management turns isolated learning into systemic improvement, and the FMEA and risk-ranking tools described earlier become the bridge between a specific failure and the anticipatory controls that prevent its recurrence elsewhere (Stamatis, 2003; ICH, 2005). 3.7 The investigation process and problem definition Method selection sits within an investigation lifecycle whose earlier stages strongly condition its success. The quality of the initial problem statement, in particular, bounds everything that follows: an imprecise or prematurely narrowed statement of the problem directs the analysis toward the wrong question regardless of the method subsequently applied. Disciplined scoping, including a clear description of what the problem is and is not, when and where it occurs, and its magnitude, is therefore a precondition of effective analysis rather than a preliminary formality (Kepner and Tregoe, 1981). Interim containment, the immediate action taken to limit harm while the investigation proceeds, must also be distinguished from corrective action, since confusing the two is a common route to premature closure (Ogbete & Aminu-Ibrahim, 2020). Evidence gathering and the construction of a factual timeline precede causal reasoning in any sound investigation. Where evidence is incomplete or is assembled only after a conclusion has been reached, the analysis risks becoming a justification rather than an inquiry. The discipline of requiring evidence for each causal link, characteristic of causal-factor charting and structured methods, guards against this failure (Gano, 2008). These process features explain why investigator competence and facilitation, identified above as organisational moderators, exert their influence: they operate largely through the quality of problem definition, evidence handling, and causal testing rather than through the mechanics of any single tool (Ogbete & Aminu-Ibrahim, 2021; Ogbete & Aminu-Ibrahim, 2022). 3.8 Common failure modes in CAPA systems The determinants can be illuminated from the opposite direction by considering the failure modes that recur in CAPA systems. These are not failures of particular tools but characteristic ways in which the investigative process breaks down, and they map directly onto the determinants. Table 5 sets out the most frequently reported failure modes alongside the determinant each violates and the corresponding remedy. Several of these failure modes are mutually reinforcing. Time pressure encourages premature closure, which favours a single linear analysis, which in turn produces a person-focused action that is easy to implement but weak in prevention, and the absence of a genuine effectiveness check allows the cycle to repeat undetected. Recognising the pattern is itself useful, because it shows that the failure modes are systemic and that addressing any one in isolation is unlikely to succeed; the remedies, like the determinants, are most effective when applied together (Peerally et al., 2017; Wu et al., 2008; Ogunwole et al., 2021). Table 5. Common failure modes in CAPA systems mapped to the determinant each violates and the corresponding remedy. Failure mode Determinant violated Remedy Premature closure on a single cause Method-problem fit; triangulation Match method to complexity; require multiple methods for serious events Defaulting to individual error and retraining Organisational climate Look for system conditions; apply a human-factors framing Imprecise or narrowed problem statement Causal accuracy Discipline scoping; specify is and is- not before analysis Conflating contributing factors with root causes Causal accuracy Require documented evidence for each causal link Action plans that are weak or unimplemented Action adequacy Design controls capable of controlling the cause; track implementation Closing actions on the assumption of success Verification of effectiveness Require objective effectiveness evidence after a defined interval Confusing interim containment with corrective action Action adequacy Distinguish containment from corrective action explicitly 3.9 A conditional model of CAPA effectiveness Drawing the strands together, the evidence does not support the intuitive proposition that some root-cause methods are simply better than others and that effectiveness follows from choosing the best one. It supports a conditional model in which methodology contributes to effectiveness only when several enabling conditions are satisfied together. The model rests on the definition of effectiveness developed earlier, which requires causal accuracy, action adequacy, and a verified result. Each determinant map onto one or more of these components. Matching method to problem and triangulating across methods serve causal accuracy, ensuring that the analysis represents the true structure of causation rather than a simplified caricature of it (Okonkwo, Agbabiaka, Ogunwole, & Mayo, 2021). Action adequacy depends on the design of the chosen control, including the quality of the procedures through which it is sustained. The human and organisational moderators condition the entire process, since neither accuracy nor adequacy can be achieved where investigators lack competence, candour, or time. Verification of effectiveness secures the final component and feeds back into the others, because a failed verification reopens the question of causal accuracy. Effectiveness emerges only when accurate analysis, adequate action, and genuine verification are all present, and the failure of any one defeats the whole. Three propositions for practice follow. First, an organisation should treat method selection as a deliberate, risk-proportionate decision supported by a defined repertoire of tools, in keeping with both the comparative literature and the regulatory proportionality principle (Doggett, 2005; ICH, 2008). Second, for complex or high-risk events it should expect multi-method triangulation rather than reliance on a single linear analysis, since the documented failures of root-cause analysis are concentrated in single-method investigations of multi-factorial events (Card, 2017; Peerally et al., 2017). Third, it should invest as heavily in the conditions and the verification surrounding analysis as in the analysis itself, because these surrounding factors are where the evidence locates the strongest leverage over recurrence (Peerally et al., 2017; Reason, 2000; Wu et al., 2008). Table 6 expresses the matching principle as a practical decision logic, offered as a heuristic consistent with the synthesized evidence rather than as a validated algorithm, and intended to be used within, not in place of, the organizational and verification conditions described above. Table 6. A risk-proportionate decision logic for matching root-cause method to problem, applied within sound organizational and verification conditions. Problem characteristics Indicated approach Rationale Simple, low-risk, apparently single cause 5 Whys, with verification Proportionate effort; linear structure adequate; verification still required Multiple plausible contributing factors Ishikawa to generate, then a narrowing method Breadth of generation followed by testing avoids premature closure Complex, high-risk or safety-critical event Fault tree or systems method, supported by other tools Combinatorial and latent causation made explicit; proportionality justifies effort Recurring problem across many events Pareto to prioritise, then targeted analysis Directs limited investigative resource to the vital few causes Process performance with measurable data DMAIC cycle Couples analysis to measurement and control to sustain the gain Forward-looking prevention FMEA Prospective scoring converts known risks into anticipatory controls 4. Discussion 4.1 Principal findings The principal finding of this review is that CAPA effectiveness in pharmaceutical quality systems is a property of a disciplined investigative system rather than of any individual root-cause technique. Methodology contributes to effectiveness only when several conditions hold together: accurate analysis through method-problem fit and, for complex events, triangulation; adequate action design, including the procedural controls that sustain it; a climate and competence that permit honest investigation; and verification that confirms the result and feeds back into the analysis. Effectiveness emerges only when these are jointly satisfied, and the failure of any one defeats the whole. This reframing has implications for practice, for regulation, and for the interpretation of the evidence base itself (Sanni, 2021a). 4.2 Implications for practice For practice, the finding redirects improvement effort. Organizations seeking to reduce deviation recurrence often invest in training staff in a single favoured tool, most commonly the 5 Whys, on the assumption that better tool use will yield better outcomes. The evidence suggests that this investment, while not wasted, addresses the weakest of the available levers. Greater returns are likely to come from developing a repertoire of methods with explicit selection criteria, from building the investigator competence and the non-punitive climate on which honest analysis depends, recognising that durable prevention requires framing human error in terms of system conditions rather than individual fault and above all from strengthening the verification of effectiveness that closes the learning loop (Peerally et al., 2017; Reason, 2000). The shift in emphasis is from the sophistication of the analytical tool to the integrity of the surrounding process. Practically, this implies that quality leaders should define which methods are available and when each is appropriate, should require triangulation for serious events, should invest in facilitator competence, and should treat the effectiveness check as a substantive evidentiary step rather than an administrative sign-off, and should ensure that procedural corrective actions are designed as durable controls rather than mere document revisions (Eyetsemitan et al., 2022; Sanni & Ajiga, 2020a). 4.3 Implications for regulation For regulation, the finding affirms the wisdom of the proportionality and effectiveness-assessment principles already embedded in the framework, while suggesting where inspectional attention is most productively directed. An organization that uses an elaborate method but never verifies the result, or that uses a single linear method for manifestly complex events, exhibits a more fundamental weakness than one whose tool documentation is imperfect. The regulatory expectation that effectiveness be assessed, not merely that actions be taken, is the requirement most closely aligned with the determinants the evidence identifies (European Commission, 2013; ICH, 2008). A regulatory focus on the quality of root-cause reasoning and the rigour of verification, rather than on the mere presence of a CAPA procedure, is therefore well supported by the synthesis, as is attention to the disciplined process coordination that effective control in a highly regulated environment demands (Eyetsemitan et al., 2020). 4.4 Limitations and future research Several limitations qualify these conclusions. The synthesis is interpretive rather than quantitative, and although this design suits a heterogeneous evidence base, it does not yield effect estimates and is subject to the selection and interpretation judgements of the reviewer. The decision to bound the literature at 2022 excludes subsequent developments. The strongest empirical sources originate outside pharmaceutical manufacturing, so the application of their findings to the pharmaceutical CAPA context rests on analogy, justified by the generic nature of the failure mechanisms but not directly demonstrated ((Sanni & Atima, 2022). Much of the methodological literature is also normative, describing intended rather than observed performance, which constrains the confidence with which causal claims about effectiveness can be made. These limitations point to a clear research agenda. The asymmetry between an extensive normative literature and a thin empirical one is the central gap. The most useful future contribution would be empirical study of CAPA effectiveness within pharmaceutical manufacturing itself, relating documented method use, investigator competence, and verification practice to subsequent recurrence, so that the conditional model proposed here could be tested rather than reasoned. Studies that link specific method-selection decisions to measured recurrence rates, and that evaluate the incremental value of triangulation and of structured effectiveness checks, would be especially valuable. A further priority is the standardization of how effectiveness is defined and reported. The heterogeneity that prevents quantitative pooling in this field is partly a heterogeneity of definitions: studies and organizations vary in what they count as recurrence, over what window, and against what baseline. Agreement on a small set of common effectiveness measures, and on minimum reporting standards for CAPA outcomes, would allow future evidence to be compared and, in time, synthesized quantitatively, and would support benchmarking that turns the currently private learning of individual quality systems into a shared evidence base (Sunday, 2019). Until such evidence exists, the present synthesis represents the best available reasoning rather than a settled empirical conclusion. 4.5 CAPA within continual improvement and knowledge management Finally, the synthesis situates CAPA within the broader purpose of the pharmaceutical quality system, which is continual improvement. The ICH Q10 model treats CAPA, change management, knowledge management, and quality risk management as interdependent enablers rather than as separate procedures (ICH, 2008). Seen in this light, an effective CAPA system is one of the principal channels through which an organisation accumulates and applies knowledge about its own processes. The verification step in particular generates knowledge about which causal hypotheses were correct, and the trending of CAPA data across events feeds the management review through which systemic improvement is steered. Treating CAPA as an isolated compliance activity, rather than as a component of an integrated learning system, is therefore itself a barrier to effectiveness, and the determinants identified in this review are best pursued as part of a deliberate strategy for organizational learning rather than as discrete procedural fixes (Carroll, 1998; ISPE, 2020). 4.6 Building investigative capability: a practical roadmap The cumulative implication of the synthesis is that improving CAPA effectiveness is primarily a matter of building investigative capability rather than of adopting a particular tool. A practical sequence follows from the determinants. An organization can begin by defining a small repertoire of methods and explicit criteria for selecting among them by problem complexity and risk, replacing the habitual default to a single tool. It can then strengthen the front end of the investigation, requiring disciplined problem definition, scoping, and evidence gathering before causal reasoning begins (Sanni, Iwuanyanwu, Essien, & Atima, 2022). For serious or recurrent events, it can mandate triangulation and independent review, and it can invest in the competence of investigators and facilitators, since the evidence locates much of the variation in outcomes there rather than in the choice of technique (Sunday et al., 2020). The capability is completed by the conditions that surround analysis. A reporting climate that protects candid disclosure, governance that resists premature closure under time pressure, and an effectiveness-verification step designed at the point of action planning and judged on objective evidence together convert sound analysis into durable prevention (Sunday, Omoegun, & Essien, 2019). None of these steps is novel, and each is already implied by the regulatory framework; the contribution of the synthesis is to show that they are interdependent and that effectiveness depends on pursuing them as a coherent system rather than piecemeal. Organisations that approach CAPA in this way are likely to find that recurrence falls not because their analytical tools have become more sophisticated, but because the conditions for using any tool well have been put in place (Peerally et al., 2017; ISPE, 2020; Yeboah, 2020). 5. Conclusion Corrective and preventive action stands at the center of the pharmaceutical quality management system, and its effectiveness depends on the quality of the root-cause analysis that precedes it. This review has shown, however, that effectiveness cannot be secured simply by selecting a superior analytical tool. The principal root-cause methodologies differ in the causal structures they can represent, and each is well suited to some problems and actively misleading for others. The decisive factors lie in the fit between method and problem, in the use of more than one method for complex events, in the human and organizational conditions that govern how analysis is conducted, and in the verification of effectiveness that confirms whether prevention was in fact achieved. The practical consequence is a change of emphasis. An organization that wishes to reduce the recurrence of deviations should treat method selection as a deliberate, risk-proportionate decision, should expect triangulation across methods for serious events, should cultivate the competence and candour on which sound investigation depends, and should make verification of effectiveness a genuine check rather than an administrative formality. Understood in this way, CAPA effectiveness is not the output of a technique but the achievement of a disciplined system that learns from its failures and confirms that it has done so. Aligning practice with that understanding offers the most credible route to fewer recurrences and to the continual improvement that the regulatory framework envisions. References Adebayo, A., Adegbite, M. P., & Ahmed, M. O. (2022). Adversarial machine learning in critical infrastructure: A conceptual framework for threat modeling AI enabled OT systems. World Journal of Innovation and Modern Technology, 6(1), 184-234. https://doi.org/10.56201/wjimt.v6.no1.2022.pg184.234 Adegbite, M. P., Adebayo, A., & Ahmed, M. O. (2020). Securing operational technology networks in electric utilities: A systematic review of NERC CIP compliance and architectural threat mitigation. International Journal of Engineering and Modern Technology, 6(3), 103-146. https://doi.org/10.56201/ijemt..pg103.146 Adegbite, M. P., Adebayo, A., & Ahmed, M. O. (2022). A security architecture model for IT and OT convergence in regulated energy networks: Design principles and governance alignment. International Journal of Computer Science and Mathematical Theory, 8(2), 81- 132. https://doi.org/10.56201/ijcsmt.v8.no2.2022.pg81.132 Adeyelu, O. O. (2022). A regulatory evidence framework for assessing unauthorized structure violations within protected airport obstacle limitation surfaces. International Journal of Scientific Research in Civil Engineering, 6(6), 248-283. https://doi.org/10.32628/IJSRCE229673 Agbabiaka, J., Okonkwo, C. S., Ogunwole, O., Mayo, W., & Okeke, O. T. (2019). Supply chain risk management model for EPC and gas processing projects. Iconic Research and Engineering Journals, 3(2), 968-980. https://doi.org/10.64388/IREV3I2-1713124 Ahiaeke Patrick, M. C., Okonkwo, C. S., Mayo, W., & Okeke, O. T. (2020). A GIS enabled framework for modern ERP procurement processes. International Journal of Multidisciplinary Research and Growth Evaluation, 1(5), 499-508. https://doi.org/10.54660/IJMRGE.2020.1.5.499-508 Ahiaeke Patrick, M. C., Okonkwo, C. S., Mayo, W., & Okeke, O. T. (2021). Model for data driven vendor evaluation and bid selection using geospatial intelligence. Shodhshauryam, International Scientific Refereed Research Journal, 4(4), 426-443. https://doi.org/10.32628/SHISRRJ214461 Ahmed, M. O., Adegbite, M. P., & Adebayo, A. (2021). Zero trust architecture for operational technology in North American critical infrastructure: A framework for implementation and resilience optimization. International Journal of Engineering and Modern Technology, 7(1), 66-113. https://doi.org/10.56201/ijemt..pg66.113 Aifuwa, S. E., Oshoba, T. O., Ogbuefi, E., Ike, P. N., Nnabueze, S. B., & Olatunde-Thorpe, J. (2020). Predictive analytics models enhancing supply chain demand forecasting accuracy and reducing inventory management inefficiencies. International Journal of Multidisciplinary Research and Growth Evaluation, 1(3), 171-181. https://doi.org/10.54660/.IJMRGE.2020.1.3.171-181 Akomolafe, and Agu. (2018). Conceptual model for enhancing audit quality through risk assessment. Akomolafe, O., & Agu, M. U. (2019a). A conceptual framework for developing risk-based internal control models in the insurance and banking sectors. Iconic Research and Engineering Journals, 2(8), 335-354. Akomolafe, O., & Agu, M. U. (2019b). A review of data-driven risk evaluation models for emerging market financial institutions. Iconic Research and Engineering Journals, 3(6), 433-448. Akomolafe, O., & Agu, M. U. (2019c). Advances in financial resilience through integrated governance and compliance strategies. Iconic Research and Engineering Journals, 2(10), 607-620. Akomolafe, O., Olaogun, B. O., Adesuyi, M. O., Ndukwe, V. U., & Sakyi, J. K. (2022). Smart contract automation model for supplier payment systems and performance benchmarking. International Journal of Multidisciplinary Research and Growth Evaluation, 3(6), 827-836. https://doi.org/10.54660/.IJMRGE.2022.3.6.827-836 Ambali, Eyetsemitan, Oyeleye, and Fadayomi. (2021). Lean Six Sigma for small enterprises: A systematic review and Lite-DMAIC adaptation framework. Ambali, K. B., Eyetsemitan, R. A., Oyeleye, A. O., & Fadayomi, O. (2021). Lean Six Sigma for small enterprises: A systematic review and Lite-DMAIC adaptation framework for resource-constrained organizations. Iconic Research and Engineering Journals, 5(5), 562- 583. https://doi.org/10.64388/IREV5I5-1716957 Aminu-Ibrahim, A. Y., Ogbete, J. C., & Ambali, K. B. (2019). Capital project delivery models for high risk healthcare infrastructure in developing national health systems. Iconic Research and Engineering Journals, 2(10), 626-649. https://doi.org/10.64388/IREV2I10-1713588 Aminu-Ibrahim, A. Y., Ogbete, J. C., & Ambali, K. B. (2020). Infrastructure driven expansion of diagnostic access across underserved and rural healthcare regions. International Journal of Multidisciplinary Research and Growth Evaluation, 1(5), 691-706. https://doi.org/10.54660/IJMRGE.2020.1.5.691-706 Andersen, and Fagerhaug. (2006). Root cause analysis: Simplified tools and techniques (2nd ed.). ASQ Quality Press. Anene, U. N., & Clement, T. (2022). A resilient logistics framework for humanitarian supply chains: Integrating predictive analytics, IoT, and localized distribution to strengthen emergency response systems. International Journal of Scientific Research in Computer Science, Engineering and Information Technology, 8(5), 398-424. Antony. (2006). Six Sigma for service processes. Business Process Management Journal, 12(2), 234-248. Arumosoye, O. M., & Obriki, O. D. (2019). Systematic review of near-miss and hazard observation data utilization in industrial safety management. Iconic Research and Engineering Journals, 3(2), 981-999. https://doi.org/10.64388/IREV3I2-1714417 Arumosoye, O. M., & Obriki, O. D. (2020). A governance-oriented conceptual model for contractor safety performance in multi-contract industrial projects. International Journal of Multidisciplinary Research and Growth Evaluation, 1(5), 728-740. https://doi.org/10.54660/.IJMRGE.2020.1.5.728-740 Arumosoye, O. M., & Obriki, O. D. (2021). Organizational learning-based conceptual maturity model for continuous safety performance improvement. International Journal of Multidisciplinary Research and Growth Evaluation, 2(1), 970-980. https://doi.org/10.54660/.IJMRGE.2021.2.1.970-980 Arumosoye, O. M., & Obriki, O. D. (2022). Conceptual risk pathway model for lifting and rigging operations in heavy industrial construction. Gyanshauryam, International Scientific Refereed Research Journal, 5(5), 346-369. https://doi.org/10.32628/GISRRJ2256237 Atakpa, M. I. (2021). A review of predictive analytics models for anti-money laundering detection in commercial banking systems. International Journal of Scientific Research in Computer Science, Engineering and Information Technology, 7(2), 778-804. https://doi.org/10.32628/CSEIT2064813 Atakpa, M. I., & Abetoh, N. F. (2022). A systematic review of machine learning advances in financial fraud detection for banking systems. International Journal of Scientific Research in Computer Science, Engineering and Information Technology, 8(2), 771-801. https://doi.org/10.32628/CSEIT23906220 Atakpa, M. I., & Abolaji, T. O. (2022). A privacy-preserving data architecture model for regulated industry analytics under GDPR and HIPAA compliance. International Journal of Scientific Research in Computer Science, Engineering and Information Technology, 8(3), 781-808. https://doi.org/10.32628/CSEIT22558 Atima, M. E., Sanni, J. O., & Attah, A. (2022). Predictive audience segmentation models resolving targeting inefficiencies in regulated professional service enterprises. Shodhshauryam, International Scientific Refereed Research Journal, 5(1), 271-303. https://doi.org/10.32628/SHISRRJ247134 Badmus, O., Dosunmu, A. A., Ozowara, D. E., & Anunagba, C. O. (2020). A comparative framework for Salesforce DevOps tooling: Evaluating Copado, Flosum, and Salesforce DX across enterprise deployment contexts. International Journal of Multidisciplinary Research and Growth Evaluation, 1(5), 1021-1031. https://doi.org/10.54660/.IJMRGE.2020.1.5.1021-1031 Basnet, A., Oghenemaiga, E., & Anene, U. N. (2021). Audience segmentation and forecasting models for enhancing targeted digital marketing effectiveness. Shodhshauryam, International Scientific Refereed Research Journal, 4(5), 279-305. Bello, K. A., Adama, A., Tawose, O. M., & Bolaji, B. O. (2022). Development and performance evaluation of a poultry bird de-feathering machine. FUOYE Journal of Engineering and Technology, 7(4). [DOI placeholder in source; verify.] Bello, K. A., Tawose, O. M., Adama, A., & Bolaji, B. O. (2022). Factor analysis of poultry birds de-feathering machines. FUOYE Journal of Engineering and Technology, 7(3). https://doi.org/10.46792/fuoyeje