Out-of-Trend Statistics in The Pharmaceutical Industry:

A Gain Leap in Assuring the Quality of The product

 

D. Mamatha1*, Hindustan Abdul Ahad2, G. Ushasree3, K. Vinod3, C. Haranath2, P. Kiran4

1Department of Pharmaceutical Quality Assurance, Raghavendra Institute of Pharmaceutical Education and Research (RIPER) - Autonomous, Ananthapuramu - 515721, Andhra Pradesh, India.

2Department of Pharmaceutics, RR College of Pharmacy, Chikkabanavara, Bengaluru - 90, Karnataka, India.

3Department of Pharmaceutical Analysis, Raghavendra Institute of Pharmaceutical Education and Research (RIPER) - Autonomous, Ananthapuramu - 515721, Andhra Pradesh, India.

4Department of Pharmacology, Raghavendra Institute of Pharmaceutical Education and Research (RIPER)-Autonomous, Ananthapuramu - 515721, Andhra Pradesh, India.

*Corresponding Author E-mail: abdulhindustan@gmail.com

 

ABSTRACT:

The primary goal of the evaluation is to ensure the product's quality by locating and managing "out of trend" (OOT) areas, utilising various techniques in the pharmaceutical sector. Regression control charts, time points, and slope control charts can all be used to identify or detect OOT. At the time of handling, OOT is divided into three categories: analytical alert, process control alert, and compliance alert. The electrical OOT Tracking Software from Ample Logic, created using low-code technology, is used to manage OOT. Identification of OOT stability results is an increasingly important topic in the pharmaceutical industry. In a perfect world, finding or detecting an OOT would be easy. However, an oversimplified system might not be sensitive enough to detect a genuine OOT. It should be chosen based on how the approach would impact the parameter being evaluated. This article outlines several tactics, such as how to recognise an unexpected single result or unusual variance. When numerous tests and time points call for OOT constraints, OOT detection can be a difficult problem. Additionally, it includes components and software that help manage OOT discoveries.

 

KEYWORDS: Error, Factors, Process, Significant, Trend.

 

 


INTRODUCTION:

A stability result that deviates from the anticipated pattern when compared to other stability batches or to past data obtained during a stability investigation is referred to as being out of trend (OOT))1-3. An out-of-trend (OOT) approach is required to reliably find and eliminate outliers from estimates for expiry and stability4,5. OOT points are considered to be unrepresentative of the test sample since they are described as being due to analytical, transcriptional, or other sorts of error.

 

 

 

If any OOT points exist, not addressing them will lead to projected rates of change that are neither representative of the therapeutic product nor its active ingredient.

 

OOT occurrences may be caused by a variety of factors. The process average may suddenly change or include an unusual data point. The standard deviation of the data could increase or decrease. The overall conclusion drawn from the data might gradually be expanded by well-known independent variables or (factors) that introduce further variety into the crucial quality features (also known as dependent variables or responses)6.

 

Keep in mind that while OOT is frequently seen as a negative event, it can sometimes be advantageous. If variability declines or the average moves in the direction of the goal, this can be an advantageous trend. A fair trend would be demonstrated by stable data that remain constant over the course of the investigation7. These changes do not necessarily need to be statistically significant for an alert to be sent and a response to be initiated8. If any OOT points exist, not addressing them will lead to projected rates of change that are neither representative of the therapeutic product nor its active ingredient. The stability study reveals the following OOT indicators9-11:

·       The amount by which something expires varies significantly over time.

·       The fit's confidence intervals are too large.

·       The points of the regression line are not lined up.

·       The presence or absence of the OOT time point significantly affects r2.

·       The residual root-mean-square error (RMSE), which is a measure of analytical error, has grown substantially.

 

OOT evaluation and removal should be utilised for the following applications and forecasting12,13:

·       Assessment of shelf-life

·       Impurity generation and trending

·       Monitoring and prediction of in-process activities

·       Storage evaluation

·       Tracking and trending of low performance

 

Regression control charts, time points, and slope control chart approaches are the strategies used to find/detect OOT14. At the time of managing them, the OOT are divided into three categories: analytical, process control, and compliance alerts15.

 

REGULATORY AND STATISTICAL APPROACHES:

According to a recent Establishment Inspection Reports (EIR) evaluation, FDA form 483s, and Warning Letters, the identification of OOT data is developing into a regulatory issue for marketed items16. Several organisations have requested the creation of standards defining how OOT stability data would be found and examined in response to the recent 483 observations17. It is crucial to note the discrepancy between OOS and OOT results. The FDA issued a draught OOS regulation as a result of a court ruling in the United States Vs. Barr Laboratories case from 199318. The subject of OOS outcomes has received substantial coverage in the academic literature and at various scientific conferences. There is no extra-legal or regulatory basis to demand a review of data within specification but not following projected trends, even though the FDA draught guideline acknowledges in a footnote that most of the advice supplied for OOS can be utilised to examine OOT outcomes19.

 

The product’s history must be considered when analysing the analytical result and defining the batch's quality, according to the 1993 court judgement in the United States Vs. Barr Laboratories20. By foreseeing the possibility of future OOS results, trend analysis provides a strong foundation for employing OOT analysis as a best practice in the industry. It also helps avert potential issues with marketed products and regulatory concerns. The extrapolation of OOT should be regulated and backed by research, just as the use of extrapolating stability data is constrained in regulatory recommendations (ICH, FDA). When an OOT data point is found, nothing more than the observation's unusualness is documented21.

 

It is vital to determine OOT stability outcomes once a batch starts functioning differently than expected using efficient and empirical statistical approaches. Before determining if a certain result is OOT, one must first establish what is expected and, in particular, what data comparisons are appropriate22.

 

The identification of OOT data can be done in one of two ways. To determine whether a batch displays the same temporal pattern as earlier time point data had revealed, the first step is to look at it23,24. The second is to assess if the batch under investigation is trending in the same direction by contrasting it with earlier batches of related products. For these two scenarios, there are different statistical methodologies. There is currently no agreement on which of these variables must be considered or whether both are equally important25.

 

Review of current and common approaches:

The authors aren't aware of any conventional statistical methodology that is frequently applied to pinpoint OOT results. However, there are a few straightforward rules that are occasionally applied; some of these methods are provided in this section26. OOT outcomes have been discovered in the past via several techniques, including the following27,28:

·       The difference between the two findings is larger than the difference between the specification and the preceding result.

·       The outcome is 5% different from the first prediction.

·       The result deviates by 3% from the previous finding.

·       The result is 5% off the average of all previous results.

·       Three results in a row that exceed some threshold.

 

These methods have the advantage of being simple to use and comprehend, and they typically don't call for distinct restrictions for each time point29.

 

The main drawback of these approaches is that they lack a statistical foundation, therefore the unpredictability of the data in a particular context affects how well they function30. This indicates that detecting a false-positive result will be more likely for parameters with high variability, but OOT results may be overlooked for parameters with low variability31.

Some of these methods also simply compare the current result to one other result. The comparison may not appropriately indicate whether the current result is OOT or not if the comparator result is erroneous (either high or low) solely by coincidence32.

 

ROOT CAUSES FOR OOT:

OOT events could specifically occur during a stability review. These typical sources and procedures are potential33-36:

·       Analyst mistakes

·       Calibration mistakes

·       Dilution errors

·       Drift in benchmarks or references.

·       Failure to follow the method's standard operating procedure and the use of nonstandard test protocols (SOP).

·       Finishing up bulk supplies.

·       Flow rate issues

·       Instrument variation

·       Process timing issues

·       Reaction time and pH inaccuracies

·       Sample component and plate flaws

·       Sample handling and selection mistakes

·       Sample preparation mistakes

·       Temperature

·       Vendor and lot variations for key reagent

 

MATERIALS AND METHODS:

The following methods are used to detect OOT37:

 

Closed loop approach:

There are 5 steps in the closed-loop approach in OOT:

 

New time points and data:

The time point should be examined for OOT potential after each new moment in time included in the stability analysis. If they meet the requirements for OOT identification - rates of change, expiry, and other factors are established. If new time points seem suspect, OOT identification, determination, and verification are used38.

 

OOT identification:

The per cent change from point-to-point as well as from the initial time point to indicate an OOT. For example, more than a 5% change from baseline may be considered a possible OOT event. There are four ways to determine whether a point is OOT39:

·       Control chart of the residuals

·       Multivariate Jack-knife distances

·       Outlier box plot of the residuals

·       Visually

 

OOT determination:

The procedure for the determination of OOT40:

·       Calculate a z score for each time point.

·       Calculate the difference (Delta) at each time point. 

·       During the data analysis, leave out and conceal the alleged OOT point.

·       Fit a linear regression line with the potential OOT time point excluded. 

·       Save the predicted response (concentration) from the linear fit.

 

The equation for the Z score is

 

                     --- (1)

 

Where Z= standard score; x= observed value; µ= mean of the sample; σ= standard deviation.

 

The most effective way to identify OOT is with a z-score with a limit. The primary distinction between this approach and others that employ z-scores is that this method evaluates the z-score with the point subtracted from it. By precisely scaling the residual error, the influence of the OOT point is removed from the residual error, allowing the OOT time frame to be evaluated appropriately based on the other measurement's errors (Muiambo et al., 2021)41.

 

OOT verification

The OOT is verified by using the following measures42:

·       Change in r˛

·       Change in RMSE

·       Change in expiry calculation

 

By comparing with or without it, one may determine the influence of the OOT time point and whether it needs to be removed. If the three measurements were just slightly altered, the OOT has not been verified and its removal is not required. When determining the stability or expiry of a pharmaceutical ingredient or product, it is generally not practical to take changes in r2, RMSE, or expiry of less than 3% into account.

 

By first including the OOT point in the stability evaluation and then removing it, the key performance criteria of the fit and the prediction are altered. Additionally, RMSE error can be contrasted with the analytical method's repeatability to ascertain whether the residual error is predominantly brought on by measurement error when performing an analysis43.

 

Stability prediction:

After the OOT point has been eliminated and confirmed, stability prediction can be done by ICH Q1E. OOT points must be plotted to demonstrate the measurement, but they must also be marked as such to indicate that they were plotted but were not included in the analysis. If it is included in the plots, it will provide full disclosure of all data at all times. Once a location has been identified as OOT, the analysis is no longer valid as new information improves the prognosis for stability44.

Regression control charts:

Along the length of the stability study, the control-chart limits encircle the regression line. This approach necessitates the presumption that data are uniformly, regularly, and independently distributed throughout all time points. For this strategy to work, all batches must have the same linear slope. To fit the data, a least-squares regression line is used. The regression line is fitted to the data for comparisons made within a batch. A regression line is fitted to the historical product database for batch comparisons, assuming a common slope but allowing for different batch intercepts. The intercepts, slope, and square root of the mean square error can all be estimated using this fit. Alternatively, a common slope estimate and standard error from the regression from historical records can be used (eq.2)45.

 

Expected result = Intercept + (slope × time) (2)

 

Calculate the predicted result (k,s), where k is a multiplier and s is the square root of the mean square error from the regression, to get the control limits for a certain time point. Stability data points within the control bounds at a given moment are in control and wouldn't be recognised as OOT. Stability data points that are OOT at a certain time point would be considered to be outside the boundaries of control and warrant additional investigation. The confidence level and, consequently, the frequency of false alarms, can be controlled based on the multiplier (k) used for the control chart. These calculations are simple to perform using a fundamental statistical software tool31.

 

Employing a prediction interval (i.e., an interval containing the future observation with a certain confidence) or tolerance interval (i.e., an interval containing a specific proportion of the future observations with a given confidence) would be a more efficient but complex strategy because these intervals reflect the number of values going into the estimate, the variation in the data, and the amount of extrapolation being performed46.

 

Depending on the circumstances, this extra intricacy might be beneficial. Keep in mind that tolerance intervals for samples with small sample sizes will likely be wide and not appropriately discriminating. Limits for the real test time could be calculated if data are not tested at the standard test times of 0, 3, 6, 9, and so on.

 

By time point method:

Using measurements taken at the same stability time point and experiences from earlier batches, the by-time-point technique is utilised to determine whether a result is within expectations. This approach makes the as sumes observations are independent and have a normal distribution.

This method calculates a tolerance interval for each stability time point using historical data. To reduce the impact of time zero discrepancies between lots, the tolerance interval might be based either on the stability results in themselves or on the variation from the lot's initial stability result.

 

One must compute the mean and standard deviation (s) for each time point to determine a tolerance interval. Finding a multiplier k using tables or by making estimations is possible. Using s, x, and K, the range can be calculated. The interval's breadth depends on the historical database's batch count and the required degrees of confidence and coverage. If the result falls outside of these boundaries at this time, it is considered out-of-the-range (OOT)47.

 

Advantages:

·       It can be used without making any assumptions about the shape of the deterioration curve when variability differs for several time points.

·       The coverage and level of assurance can be tailored to a given product's requirements.

 

Challenge:

When current data are not checked at the nominal time points, this approach has a problem. Then, as a rough guide, limitations calculated for the before or subsequent nominal time point may be employed. It depends on the pace of change whether this estimate is appropriate and whether to choose boundaries for the prior or subsequent time point48.

 

Slope control chart method:

Making a control chart for the slope at each time point is another method for identifying OOT outcomes for single results. Using this strategy, batches can be compared. For each instant in time, a least squares regression is fitted, taking into account all data up to that point. Because the slope is normally distributed, the slope estimate for each batch is used to determine an overall slope estimate and control limits. The tolerance interval, where “x” and “s” are the means and standard deviation of the historical slope estimates OOT limits generated for the slopes at each time point, and k is chosen to ensure the required coverage.

 

Advantage:

·       The bounds are wider at earlier time points because the slope estimate is more variable when fewer data points are included in the regression

·       This method compares slopes to see if all batches behave the same.

 

Disadvantage:

A slope calculation must be done to determine whether a data point is OOT. Often, this calculation is not performed automatically after each time point, and the accountable analyst might not have easy access to prior data. The restrictions might not be suitable if data are not checked at the nominal time points. However, there shouldn't be any significant effects if there are little variations between the nominal age and the test age49.

 

HANDLING OF OOT RESULTS:

Three categories are used to classify out-of-trend findings (table 1)50:

 

Table 1: Classification of OOT based on alert type

Stage

Name of the alert

Description

1

Analytical

If only one outcome, such as an out-of-trend assay value, is abnormal but falls within specified bounds.

2

Process control

An atypical sample is suggested by a string of trustworthy data points or a string of finished goods batches.

3

Compliance

A compliance alert occurs when a selected stability study or release result does not meet specifications throughout the expiration time or with the regulatory specification of the same product.

 

The Quality unit (QU) should include all production-related disciplines, including engineering, manufacturing, method improvement, and maintenance, before taking the lead on the investigations. After being recognised, any issues need to be thoroughly investigated to prevent future problems. All statistics and files pertinent to the production system must go through a thorough and completely documented review to establish the potential application of the OOT result. The evaluation report must include the following51:

·       A concise explanation of the many production components that might be at fault for the problem.

·       A detailed basis for the investigation.

·       The results of the documentation review to determine whether this is an ongoing issue and to identify the real or likely cause.

·       Documenting the corrective actions.

 

OOT software for handling:

The electrical OOT Tracking Software from Ample Logic was created utilising low-code technology. According to the needs of the organisation, it can keep and retrieve any relevant records. The EU Annex 11, the GMP standard, and other international and national production standards are all ensured by our Out of Trend programme38,39.

 

CONCLUSION:

Identification of OOT stability results is an increasingly important topic in the pharmaceutical industry. In a perfect world, finding or detecting an OOT would be easy. However, an oversimplified system might not be sensitive enough to detect a genuine OOT. It should be chosen based on how the approach would impact the parameter being evaluated. This article outlines several tactics, such as how to recognise an unexpected single result or unusual variance. When numerous tests and time points call for OOT constraints, OOT detection can be a difficult problem. Additionally, it includes components and software that help manage OOT discoveries.

 

CONFLICT OF INTEREST:

The authors have no conflicts of interest regarding this investigation.

 

REFERENCES:

1.      Mihalovits M, Kemény S. Regression control chart with unknown parameters for detection of out-of-trend results in pharmaceutical on-going stability studies. Journal of Pharmaceutical and Biomedical Analysis. 2020; 188:113375.

2.      Szél E, Bozó R, Hunyadi-Gulyás É, Manczinger M, Szabó K, Kemény L, Bata-Csörgő Z, Groma G. Comprehensive proteomic analysis reveals the intermediate stage of non-lesional psoriatic skin and points out the importance of proteins outside this trend. Scientific Reports. 2019; 9(1):1-5.

3.      Mohammed Imtiaz Ahmed. Web Content Management in Library Science: An Information Technology Perspective. Asian J. Management. 2010; 1(2): 84-89.

4.      Mehvish Anjum. Strengthening Financial Position through Profit Maximisation a Study on Controllable Factors via Indian Wrist Watch Industry. Asian J. Management. 2011; 2(2): 77-80.

5.      Alhakeem MA, Ghica MV, Pîrvu CD, Anuța V, Popa L. Analytical quality by design with the lifecycle approach: A modern epitome for analytical method development. Acta Marisiensis-Seria Medica. 2019; 65(2): 37-44.

6.      Ahad HA, Haranath C, Rahul Raghav D, Gowthami M, Naga Jyothi V, Sravanthi P. Overview on recent optimization techniques in gastro retentive microcapsules by factorial design. Int J Pharm Sci Res. 2019; 10(9): 247-54.

7.      Chinthaginjala H, Ahad HA, Bhargav E, Pradeepkumar B. Central Composite Design Aided Formulation Development and Optimization of Clarythromycin Extended-Release Tablets. Indian Journal of Pharmaceutical Education and Research. 2021; 55(2): 395-406.

8.      Bianchi M, Cavaliere A, Chiaroni D, Frattini F, Chiesa V. Organisational modes for Open Innovation in the bio-pharmaceutical industry: An exploratory analysis. Technovation. 2011; 31(1): 22-33.

9.      Schuhmacher A, Germann PG, Trill H, Gassmann O. Models for open innovation in the pharmaceutical industry. Drug Discovery Today. 2013; 18(23-24): 1133-7.

10.   Jaiswal H, Muddukrishna BS, Kulyadi GP. Data integrity violations: A challenge to the pharmaceutical industry. Quality Assurance. 2020; 5: 16.

11.   Chinthaginjala H, Ahad HA, Samhitharao B, Yasaswini VS, Madana G, Varam NJ, Dasari R. Formulation and In Vitro Evaluation of Fast Disintegrating Tablets of an Anti-Inflammatory Drug using Natural Super Disintegrating Agents. Int. J. Life Sci. Pharma Res. 2021; 2: 9-18.

12.   Chaitanya MV, Chinni S, Jyothia MV, Reddy YP. The Importance of Quality Management System for a Successful Health Care Industry: A Review Based on Case Studies. Journal of Chemical and Pharmaceutical Research. 2018; 10(8): 143-9.

13.   Shravani Y, Ahad HA, Haranath C, gari Poojitha B, Rahamathulla S, Rupasree A. Past Decade Work Done on Cubosomes Using Factorial Design: A Fast Track Information for Researchers. Int. J. Life Sci. Pharma Res. 2021; 11(1): 124-135.

14.   Flanders W. Opting out: Enrollment trends in response to continued public school shutdowns. Journal of School Choice. 2021; 15(3): 331-43.

15.   S. Muneer, Hindustan Abdul Ahad, Chandra Sekhar Kothapalli Bonnoth. A Novel Stability Indicating RP-HPLC Assay Method Development and Validation for the Quantification of Cyamemazine Tartrate in bulk and its Pharmaceutical Dosage Form. Asian J. Pharm. Ana. 2018; 8(3): 169-173.

16.   Yu B, Zeng L, Ren P, Yang H. A unified framework for detecting out-of-trend results in stability studies. Statistics in Biopharmaceutical Research. 2018; 10(4):237-43.

17.   Hindustan AA, Babu UA, Nagesh K, Kiran DS, Madhavi KB. Fabrication of glimepiride Datura stramonium leaves mucilage and poly vinyl pyrrolidone sustained release matrix tablets: in vitro evaluation. Kathmandu University Journal of Science, Engineering and Technology. 2012; 8(1): 63-72.

18.   Hindustan Abdul Ahad, Chinthaginjala Haranath, Bhupalam Pradeep Kumar, Debaraj Roy, Bhavani Haribabu Sai Dharani, Momin Umaira Ayisha. Asian Journal of Research in Pharmaceutical Sciences. 2021; 11(3): 247-2.

19.   Wande DP, Sangeda RZ, Tibalinda P, Mutta IK, Mkumbwa S, Bitegeko A, Kaale E. Pharmaceuticals imports in Tanzania: an overview of private sector market size, share, growth and projected trends to 2021. PloS one. 2019; 14(8): e0220701.

20.   Yadiki MN, Suggala VS, Puchalapalli DS, Ahad HA. Temperature and Exposure Time Impact on the Extraction of Opuntiaficus-indica and Opuntia dillenii Cladodes on% Yield as a Response: Screening using Design Expert Software. Global Journal of Medical, Pharmaceutical, and Biomedical Update. 2022; 17.

21.   Darmawan LV, Hutajulu M, Sianturi JV, Muda I. Fairness, Disclosures and Future Trends in Accounting for Pharmaceutical Companies. Journal of Pharmaceutical Negative Results. 2022: 1638-44.

22.   Bandagisa Fouziya, Hindustan Abdul Ahad, Dontha Swamy Charan, Jagarlamudi Sri Vidya, Ummadi Chandana Reddy, Puli Nandini Reddy. Fabrication and Evaluation of Cefpodoxime Proxetil Niosomes. Asian Journal of Pharmacy and Technology. 2022; 12(2):109-2.

23.   Hindustan Abdul Ahad, Haranath Chinthaginjala, Gangireddy Jayasimha Reddy, Pasupuleti Dheeraj Krishna, Syed Rahamathulla, Jyothi Vinay Krishna. Abbreviated New Drug Application Process: A Gift for The Industry and The Patients. Asian Journal of Research in Chemistry. 2021; 14(5):331-5.

24.   Chinthaginjala H, Barghav GC, Reddy CM, Pradeepkumar B, Ahad HA, Akbari BV, Dholakiya RB, Shiyani BG, Lodhiya DJ, Shastry CS. Formulation and in vitro evaluation of floating tablets of dicloxacillin sodium using different polymers. Journal of Young Pharmacists. 2019; 11(3): 247.

25.   Sahani P, Srilatha KS, Ahad HA, Sah SK, Dkhar B, Prajapati RK. From Concept to Assessment: Creating an Oral In-Situ Gelling System with Sucralfate. KEPES. 2023; 21(3): 555-67.

26.   Uddin M. Blockchain Medledger: Hyperledger fabric enabled drug traceability system for counterfeit drugs in pharmaceutical industry. International Journal of Pharmaceutics. 2021; 597: 120235.

27.   Suresh K, Hindustan AA, Satyanarayana SV. Antioxidant activity and Hepatoprotective potential of Ethanolic leaf extract of Artabotryshexapetalus against various Hepatotoxins induced Hepatotoxicity in Albino wister Rats. Int. J. Res. Pharm. Sci. 2021; 12: 1679-88.

28.   Hindustan Abdul Ahad, Haranath Chintaginjala, Syed Rahamathulla, AswarthanarayanaRupasree, Anegondithimmappa Sajan Kumar, Barji Prasanna Pallavi. Pathfinder Nanosponges for Drug Targeting by Factorial Design: A Glance Review. Research Journal of Pharmaceutical Dosage Forms and Technology. 2021; 13(4):341-4.

29.   Kousar S, Ahad HA, Chinthaginjala H, Babafakruddin P, Lakunde J, Tarun K. Gas Generating Floating Tablets: A Quick Literature Review for the Scholars. Asian Journal of Research in Chemistry. 2022 ;15(2): 171-5.

30.   Abualhaija S, Arora C, Sabetzadeh M, Briand LC, Vaz E. A machine learning-based approach for demarcating requirements in textual specifications. In2019 IEEE 27th International Requirements Engineering Conference (RE) 2019 Sep 23 (pp. 51-62). IEEE.

31.   Babu GN, Muthukaruppan M, Ahad HA. Impact of Azadirachtaindica Fruit Mucilage on particle size and swelling index in Central Composite Designed Acyclovir mucoadhesive microspheres. Baghdad Science Journal. 2023;20(2): 425.

32.   Kappala Sailaja, Hindustan Abdul Ahad, Haranath Chinthaginjala, Renuka Gudisipalli, SugaliIndravath Rajyalakshmi, Yamuna Vagganagari. Research Journal of Pharmaceutical Dosage Forms and Technology. 2022; 14(3): 245-8.

33.   Garcia-Muińa FE, González-Sánchez R, Ferrari AM, Volpi L, Pini M, Siligardi C, Settembre-Blundo D. Identifying the equilibrium point between sustainability goals and circular economy practices in an Industry 4.0 manufacturing context using eco-design. Social Sciences. 2019; 8(8): 241.

34.   Molodtsova T, Papell DH. Out-of-sample exchange rate predictability with Taylor rule fundamentals. Journal of International Economics. 2009; 77(2): 167-80.

35.   Sandell D, Tougas T. Quality considerations in the establishment of specifications for pharmaceuticals. Statistics in Biopharmaceutical Research. 2012; 4(2): 125-35.

36.   Wu J, Song X, Xiao Z. Testing for Trend Specifications in Panel Data Models. Journal of Business & Economic Statistics. 2022:1-4.

37.   Hindustan Abdul Ahad, Haranath Chinthaginjala, Nagarajugari Kavyasree, Jyothi Vinay Krishna, Sushma Papireddypalli, Kotlo Saijyoshna. A Recent attempt made on Fast Disintegrating Tablets: A Precise Review. Research Journal of Pharmaceutical Dosage Forms and Technology. 2022; 14(3):229-2.

38.   Industan Abdul Ahad, Haranath Chinthaginjala, Nagarajugari Kavyasree, Jyothi Vinay Krishna, Sushma Papireddypalli, Kotlo Saijyoshna. A Recent attempt made on Fast Disintegrating Tablets: A Precise Review. Research Journal of Pharmaceutical Dosage Forms and Technology. 2022; 14(3): 229-2.

39.   Shobhit Bajpayee, Swati Tiwari. Recent Trend of Retail Business in Chhattisgarh: With special reference to Bilaspur, Chhattisgarh. Asian J. Management. 2014; 5(2): 261-264.

40.   Luci-Greulich A, Thévenon O. The Impact of Family Policies on Fertility Trends in Developed Countries: L’influence des politiques familiales sur les tendances de la fécondité des pays développés. European Journal of Population/Revue européenne de Démographie. 2013; 29: 387-416.

41.   Muiambo CC, Joao IM, Navas HV. Lean waste assessment in a laboratory for training chemical analysts for the pharmaceutical industry. International Journal of Lean Six Sigma. 2022; 13(1): 178-202.

42.   Babu GN, Muthukarupan M, Ahad HA, Sreedhar V. Fabrication and Preliminary Assessment of Neem Fruit Mucilage as Mucoadhesive Abetting Assets with Methpol-934P for Acyclovir Delivery from Mucoadhesive Microcapsules. Biomedical and Pharmacology Journal. 2022; 15(4): 2179-84.

43.   Roesti D, Goverde M. Data Integrity and Microbiological Excursion Handling. Pharmaceutical Microbiological Quality Assurance and Control: Practical Guide for Non‐Sterile Manufacturing. 2019: 401-28.

44.   Jain AA, Mane-Kolpe PD, Parekar PB, Todkari AV, Sul KT, Shivpuje SS. Brief review on Total Quality Management in Pharmaceutical Industries. International Journal of Pharmaceutical Research and Applications. 2022; 7(5):1030-6.

45.   Santos RN, Yamouni S, Albiero B, Vicente R, A Silva J, FB Souza T, CM Freitas Souza M, Lei Z. Gradient boosting and Shapley additive explanations for fraud detection in electricity distribution grids. International Transactions on Electrical Energy Systems. 2021; 31(9): e13046.

46.   Neeti Roy, Yogendra Jain. Education through ICT Technology: Need of an hour. Asian J. Management. 2014; 5(2): 246-249.

47.   Patel Dhaval V, Patel Biraju D, Patel Nilesh K, Sheth Navin R, Dabhi Mahesh R, Dudhrejiya Ashvin V. Quality Improvement Methodologies in Pharmaceutical Manufacturing. Asian J. Management. 2010; 1(1): 1-3

48.   Xia Y, He L, Li Y, Fu Y, Xu Y. A dynamic credit scoring model based on survival gradient boosting decision tree approach. Technological and Economic Development of Economy. 2021; 27(1): 96-119.

49.   Stanisic N, Radojevic T, Stanic N. Predicting the type of auditor opinion: Statistics, machine learning, or a combination of the two?. Machine Learning, or a Combination of the Two. 2019: 1-58.

50.   Cadinoska MV, Popstefanova N, Ilievska M, Karadzinska E, Jovanoska MD, Dodov MG. Trending and Out-of-Trend results in pharmaceutical industry. Macedonian Pharmaceutical Bulletin. 2019; 65(1).

51.   Modhave D, Barrios B, Paudel A. PVP-H2O2 complex as a new stressor for the accelerated oxidation study of pharmaceutical solids. Pharmaceutics. 2019; 11(9): 457.

52.   Tome T, Žigart N, Časar Z, Obreza A. Development and optimization of liquid chromatography analytical methods by using AQbD principles: Overview and recent advances. Organic Process Research & Development. 2019; 23(9): 1784-802.

 

 

Received on 14.02.2023                    Modified on 17.06.2023

Accepted on 27.09.2023                   ©AJRC All right reserved

Asian J. Research Chem. 2023; 16(6):423-428.

DOI: 10.52711/0974-4150.2023.00069