Page 1160
www.rsisinternational.org
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue II, February 2026
Linking Demand Forecasting to Operational Outcomes: A Cross-
Sectional Supply Chain Analysis of Kenyan Public Hospitals
Abuya, Joshua Olang’o
1
, Okello, Sharone Adhiambo
2
1
School of Business & Economics, Kibabii University, Kenya
2
School of Business & Economics, Jaramogi Oginga Odinga University of Science and Technology, Kenya
DOI: https://doi.org/10.51583/IJLTEMAS.2026.15020000102
Received: 26 February 2026; Accepted: 03 March 2026; Published: 20 March 2026
ABSTRACT
Operational performance within Kenyan public hospitals has become an issue of growing public and policy
concern, particularly in the context of frequent medicine stock imbalances and service delivery inefficiencies.
Persistent challenges in balancing drug overstocking and stock-outs point to weaknesses in demand planning
within hospital supply chains. This study examines the relationship between demand forecasting practices and
operational outcomes in public hospitals in Kenya, using evidence from Siaya County. The research is anchored
on the Resource-Based View (RBV) and Network Perspective Theory to explain how internal forecasting
capabilities and inter-organizational supply chain relationships influence hospital performance. A cross-sectional
survey design was employed involving personnel drawn from procurement, pharmacy, stores, and administrative
departments across six public hospitals. Data were analyzed using descriptive statistics, correlation analysis, and
linear regression modelling. The findings reveal a strong and statistically significant positive relationship
between demand forecasting practices and operational outcomes (β = 0.876, p < 0.05). The model explains
approximately 70.1% of the variation in operational performance (R² = 0.701). These results suggest that
hospitals that institutionalize structured forecasting practices within their supply chain systems achieve improved
operational efficiency and enhanced service delivery outcomes. The study concludes that strengthening
forecasting capabilities is critical for improving drug availability and reducing service disruptions in public
healthcare systems. The paper recommends adoption of data-driven forecasting approaches, strengthened supply
chain coordination, and integration of digital health logistics systems to enhance healthcare service delivery in
Kenya.
Keywords: Demand Forecasting, Healthcare Supply Chain, Operational Performance, Public Hospitals.
INTRODUCTION
concern, particularly in the context of frequent medicine stock imbalances and service delivery inefficiencies.
Persistent challenges in balancing drug overstocking and stock-outs point to weaknesses in demand planning
within hospital supply chains. This study examines the relationship between demand forecasting practices and
operational outcomes in public hospitals in Kenya, using evidence from Siaya County. The research is anchored
on the Resource-Based View (RBV) and Network Perspective Theory to explain how internal forecasting
capabilities and inter-organizational supply chain relationships influence hospital performance. A cross-sectional
survey design was employed involving personnel drawn from procurement, pharmacy, stores, and administrative
departments across six public hospitals. Data were analyzed using descriptive statistics, correlation analysis, and
linear regression modelling. The findings reveal a strong and statistically significant positive relationship
between demand forecasting practices and operational outcomes (β = 0.876, p < 0.05). The model explains
approximately 70.1% of the variation in operational performance (R² = 0.701). These results suggest that
hospitals that institutionalize structured forecasting practices within their supply chain systems achieve improved
operational efficiency and enhanced service delivery outcomes. The study concludes that strengthening
forecasting capabilities is critical for improving drug availability and reducing service disruptions in public
healthcare systems. The paper recommends adoption of data-driven forecasting approaches, strengthened supply
Page 1161
www.rsisinternational.org
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue II, February 2026
chain coordination, and integration of digital health logistics systems to enhance healthcare service delivery in
Kenya.
Keywords: Demand Forecasting, Healthcare Supply Chain, Operational Performance, Public Hospitals.
INTRODUCTION
Across the world, healthcare institutions, both public and private, prioritize high levels of operational efficiency
and effective service delivery as core performance indicators (de Vries & Huijsman, 2011). Public hospitals, in
particular, are mandated to provide continuous, 24-hour services, which requires accurate planning and timely
availability of essential medicines and medical supplies. In this context, demand forecasting plays a critical role
in anticipating patient needs, guiding procurement decisions, and ensuring uninterrupted healthcare delivery.
Where forecasting systems are weak or inaccurate, hospitals often experience stock imbalances that disrupt
service provision and compromise operational outcomes (Stevenson, 2010).
In Kenya, public hospitals face persistent challenges in predicting patient demand patterns and determining
appropriate stock levels to meet fluctuating healthcare needs (Toroitich et al., 2022). The difficulty of balancing
shortages and excess inventory reflects gaps in supply chain forecasting and planning mechanisms. Ineffective
demand forecasting can result in stock-outs, delayed treatments, increased patient waiting times, and overall
inefficiencies in hospital operations. Consequently, operational outcomes, such as service reliability,
responsiveness, and cost efficiency, are directly affected.
Evidence from national policy and health-system assessments indicates that essential-medicine availability
constraints remain a significant barrier to effective service delivery, and stock-outs can contribute to avoidable
morbidity and mortality (Ministry of Health [Kenya], 2014; Toroitich et al., 2022). In settings where patients are
frequently referred to private pharmacies due to public-facility stock-outs, out-of-pocket costs can deter medicine
access and undermine treatment continuity (Toroitich et al., 2022). These challenges highlight systemic
weaknesses in forecasting and supply chain planning within Kenyan public hospitals. Against this backdrop, the
present study adopts a cross-sectional supply chain perspective to examine how demand forecasting practices
are linked to operational outcomes in Kenyan public hospitals.
Hyndman and Athanasopoulos (2021) describe demand forecasting as the process of estimating anticipated
demand within a defined future timeframe. In a similar vein, Lysons and Farrington (2016) explain it as the
projection of future production or distribution requirements based on historical data and other variables that may
influence organizational outcomes. Demand forecasting is commonly operationalized using quantitative and
qualitative approaches, including time-series methods that extrapolate historical patterns and causal
(explanatory) methods that model relationships between predictors and the forecast variable (Chopra & Meindl,
2021; Hyndman & Athanasopoulos, 2021; Lysons & Farrington, 2016). Within supply chain management,
effective forecasting supports inventory strategies such as just-in-time systems, where materials and components
are supplied precisely when required in the production or service delivery process, thereby minimizing holding
costs and stock imbalances (Chopra & Meindl, 2021).
Operational performance, on the other hand, refers to how well an organization performs relative to established
benchmarks and standards. Neely (2005) defines it as an organization’s performance measured against
predefined criteria including regulatory compliance, waste minimization, and productivity. He further
emphasizes that operational performance encompasses indicators of efficiency and effectiveness, forming a
framework for assessing how well organizational activities achieve intended objectives. According to Neely
(2005), key dimensions of operational performance include efficiency, effectiveness, quality, timeliness,
flexibility, cost management, and productivity.
Kaplan and Norton (1992), through the Balanced Scorecard framework, propose that organizational performance
should be assessed from four complementary perspectives: financial performance, customer outcomes, internal
business processes, and learning and growth. This multidimensional approach ensures a balanced evaluation by
integrating short- and long-term goals, financial and non-financial indicators, as well as internal efficiencies and
external stakeholder outcomes. Within public healthcare systems, such a framework is particularly relevant for
Page 1162
www.rsisinternational.org
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue II, February 2026
examining how supply chain capabilities, such as demand forecasting, translate into measurable operational
outcomes.
Stevenson (2009) notes that performance concerns within supply chain and inventory systems typically revolve
around two key issues: service levels and cost control. From a service perspective, organizations must ensure
that the right products are available in the correct quantities, at the appropriate location, and at the required time.
From a cost perspective, attention is directed toward minimizing ordering and holding costs. In hospital settings,
inaccurate demand forecasting can compromise both dimensions, resulting in stock-outs that affect patient care
or excess inventory that strains limited budgets.
Service quality measurement is therefore central to operational success in healthcare institutions (Asubonteng et
al., 1996; Gefen, 2002; Parasuraman et al., 1988). Gefen (2002) conceptualizes service quality as the gap
between patients’ expectations and their actual service experience, while Asubonteng et al. (1996) similarly
define it as the difference between anticipated and perceived service performance. In public hospitals, this gap
is often influenced by the reliability of forecasting systems that determine drug availability, staffing levels, and
resource allocation.
Operational outcomes in healthcare can be evaluated using indicators such as efficiency, effectiveness, regulatory
compliance, cycle time, productivity, and waste reduction (Abdel-Maksoud et al., 2008; Neely, 2005). Improving
operational efficiency requires monitoring both input variables (e.g., medical supplies, staff time) and output
measures (e.g., number of patients served, treatment outcomes) (Abdel-Maksoud et al., 2008). In public hospitals,
efficiency may be reflected in reduced patient waiting times, optimal bed occupancy rates, shorter lengths of
stay, timely admissions and discharges, reduced mortality rates, and improved coordination across departments.
Empirical studies in Kenya and comparable settings associate operational performance in health facilities with
essential medicine availability, timeliness of service delivery, reduced lead times, and patient satisfaction
(Toroitich et al., 2022). These indicators underscore the importance of robust supply chain planning mechanisms.
This study was grounded in two complementary theoretical lenses: The Resource-Based View (RBV) and the
Network Perspective Theory, both of which provide a foundation for understanding how demand forecasting
capabilities relate to operational outcomes in public hospitals. The Resource-Based View, originally advanced
by Edith Penrose (1959), conceptualizes the organization as a bundle of tangible and intangible resources
structured to achieve productive purposes. Wernerfelt (1984) further developed this perspective by arguing that
firms derive superior performance from valuable, rare, inimitable, and well-organized resources and capabilities.
RBV therefore emphasizes that sustainable performance improvements arise not merely from possessing
resources, but from effectively deploying them through coordinated processes and routines (Ray et al., 2004). In
the context of Kenyan public hospitals, demand forecasting systems, data analytics capabilities, skilled
personnel, and integrated information systems can be viewed as strategic resources. When embedded within
supply chain processes, these capabilities enhance service reliability, cost control, and overall operational
efficiency.
Complementing RBV, the Network Perspective Theory, traced to Jacob Moreno’s (1930) early work on
sociograms, emphasizes the importance of relationships among interconnected actors. The theory examines how
nodes (individuals, groups, or organizations) are linked through various ties, including communication, formal
authority, trust-based relationships, workflow exchanges, and resource flows (Wasserman & Faust, 1994). In
supply chain contexts, these ties extend beyond interpersonal relations to include interactions between hospitals,
suppliers, government agencies, and regulatory bodies. Effective demand forecasting in public hospitals depends
not only on internal capabilities but also on timely information sharing, supplier coordination, and collaborative
planning across the healthcare network (Borgatti & Li, 2009).
A key criticism of the Network Perspective Theory is that, although it strongly emphasizes collaboration and
knowledge exchange, it offers limited explanation regarding how networks are initiated, structured, or sustained
over time. The theory has often been applied in large organizations with complex stakeholder arrangements,
potentially limiting its applicability in smaller or less formalized institutional contexts. Additionally, certain
informal or tacit dimensions of knowledge sharing may not be fully captured through network analysis
Page 1163
www.rsisinternational.org
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue II, February 2026
frameworks (Bosua & Scheepers, 2007). Despite these limitations, the theory remains relevant to this study as it
helps explain the link between demand forecasting and operational outcomes. In Kenyan public hospitals,
effective forecasting depends on coordinated interactions among procurement units, pharmacists, clinicians,
suppliers, and government agencies. These interconnected actors form a healthcare supply network whose
collaborative functioning enhances forecasting accuracy and responsiveness within the hospital system.
Empirical evidence also underscores the importance of structured inventory practices in improving
organizational outcomes. For example, inventory classification approaches(e.g., ABC analysis) have long been
shown to improve inventory control and operational decision-making (Flores & Whybark, 1986).
Kalchschmidt (2014) investigated demand forecasting practices among manufacturing firms in Italy and
established that forecasting is a critical mechanism for aligning production planning with broader supply chain
activities. The study employed a descriptive research design with simple random sampling and utilized both
descriptive and inferential statistical methods in data analysis. Findings revealed that many firms adopted
Material Requirements Planning (MRP) systems and Economic Order Quantity (EOQ) models to structure their
forecasting processes. Operational performance was evaluated mainly in terms of cost efficiency and delivery
reliability, and the results demonstrated that firms embedding forecasting within their inventory systems achieved
measurable improvements in performance. Despite these contributions, the study was largely confined to a
manufacturing context characterized by relatively stable production cycles and predictable demand patterns.
Such conditions differ significantly from the healthcare environment, where demand is often uncertain,
emergency-driven, and influenced by epidemiological trends.
Evidence from health systems research indicates that supply chain weaknesses, including inadequate forecasting
and inventory practices, are associated with essential medicine stock-outs and compromised service continuity
(Leung et al., 2016; Toroitich et al., 2022).
The Kenya Health Policy emphasizes the government’s commitment through the Ministry of Health to
collaborate with public health institutions to ensure timely delivery of medical goods and services as a foundation
for quality healthcare (Ministry of Health [Kenya], 2014). Similarly, the Constitution of Kenya (2010) mandates
the Ministry of Health to formulate policies, set standards, regulate service provision, and create an enabling
framework for effective healthcare delivery. Under Kenya’s devolved governance structure, county governments
are thus responsible for managing county health services, including pharmacies, ambulance services, primary
healthcare promotion, food safety licensing, and public health functions such as waste management.
At the operational level, sub-county hospitals prepare annual procurement plans, which are consolidated at the
county level. Major pharmaceutical supplies are typically sourced through the Kenya Medical Supplies Authority
(KEMSA), while minor items may be procured locally through quotation processes when urgent needs arise.
Kenya’s public health system continues to face pressures related to access and affordability, with poverty
dynamics shaping reliance on public facilities (Kenya National Bureau of Statistics, 2021).
Although universal healthcare was prioritized under the government’s Big Four Agenda, persistent medicine
availability challenges continue to undermine service delivery in many counties (Toroitich et al., 2022; Walukana
et al., 2021). Ensuring uninterrupted drug availability requires robust demand forecasting and coordinated supply
chain planning rather than reliance on reactive procurement practices.
Despite significant investments by Siaya County, in collaboration with the World Health Organization, in areas
such as service delivery, health governance, infrastructure, and information systems, limited emphasis has been
placed on strengthening forecasting-driven supply chain systems. Ministry of Health reports (2025) further
indicate that Siaya County has recorded high mortality rates, with malaria and HIV/AIDS among leading causes
of death, often exacerbated by medicine unavailability. This context underscores the critical need to examine
how improved demand forecasting within hospital supply chains can enhance drug availability, strengthen
service delivery, and ultimately improve operational outcomes in Kenya’s public healthcare system.
Page 1164
www.rsisinternational.org
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue II, February 2026
Statement of the Problem
Public hospitals are mandated to operate continuously, providing round-the-clock healthcare services. To fulfill
this responsibility effectively, health facilities must maintain adequate and timely availability of essential
medicines and medical supplies. However, inefficiencies in pharmaceutical inventory management remain a
persistent challenge in many public healthcare systems, particularly in low- and middle-income countries. Weak
demand forecasting practices often lead to imbalances between supply and patient demand, resulting in either
overstocking or frequent stock-outs of essential medicines. Such supply chain inefficiencies disrupt healthcare
delivery, increase treatment delays, and ultimately compromise operational performance within health facilities
(de Vries & Huijsman, 2011; Leung et al., 2016). In Kenya, medicine availability challenges continue to
undermine the effectiveness of public healthcare delivery despite ongoing policy reforms aimed at improving
access to healthcare services. Studies examining pharmaceutical supply chains in Kenya have reported that
medicine stock-outs remain common across public health facilities, often forcing patients to seek medications
from private pharmacies where costs may be prohibitive (Toroitich et al., 2022). These supply disruptions are
frequently linked to weaknesses in forecasting systems, procurement planning, and coordination across
healthcare supply chain actors. The situation is particularly pronounced in counties with high disease burdens
such as Siaya County. Located along Lake Victoria, Siaya experiences relatively high prevalence rates of malaria
and HIV/AIDS, which place significant pressure on the county’s public health facilities. Ensuring reliable
availability of essential medicines in such settings requires effective forecasting systems capable of anticipating
fluctuating disease patterns and patient demand. However, inconsistent forecasting practices within hospital
supply chains often result in unpredictable drug availability and service delivery challenges. These operational
challenges have important implications for Kenya’s ongoing efforts to achieve Universal Health Coverage
(UHC), which seeks to ensure that all individuals have access to essential health services without financial
hardship. Evidence from Kenya’s UHC pilot implementation indicates that although healthcare access has
improved, persistent shortages of medical supplies and medicines continue to constrain service delivery within
public health facilities (Walukana et al., 2021). Anchored on the Resource-Based View and Network Perspective
Theory, the study investigates how forecasting capabilities and supply chain coordination mechanisms contribute
to improved operational efficiency and service delivery within public healthcare institutions. Public hospitals are
mandated to operate continuously, providing round-the-clock healthcare services. To fulfill this responsibility
effectively, they must maintain adequate levels of essential medicines and medical supplies. Inefficiencies in
stock management have been widely associated with disruptions in healthcare delivery and diminished service
quality (Leung et al., 2016; Toroitich et al., 2022). A persistent challenge for public hospitals is determining
optimal inventory levels that can meet patient demand without resulting in either excess stock or frequent
shortages. Achieving this balance between overstocking and stock-outs remains a significant operational concern
(Leung et al., 2016). Evidence from Kenyan health-system research indicates that public facilities can experience
shortages of critical items, with patients frequently referred to private pharmacies due to stock-outs, and
affordability constraints limiting medicine uptake (Toroitich et al., 2022). Against this backdrop, and aligned
with the broader objective of linking demand planning to service outcomes, the present study sought to examine
the effect of demand forecasting practices on operational performance in public hospitals in Siaya County,
Kenya.
METHODOLOGY
Research Design
The study employed a cross-sectional survey research design. According to Cooper and Schindler (2014), a
cross-sectional survey involves collecting data at a single point in time to describe and examine the prevailing
conditions of a given population. This design was considered appropriate because it enabled the researcher to
obtain a comprehensive snapshot of existing demand forecasting practices and operational outcomes across
public hospitals in Siaya County. A cross-sectional approach is particularly suitable for gathering primary data
from a relatively large population efficiently by selecting a representative sample. It allows for the assessment
of attitudes, practices, and institutional processes without the need for prolonged follow-up. By capturing data
concurrently from multiple facilities, the design facilitated comparative analysis and strengthened the reliability
of findings regarding prevailing supply chain practices. In alignment with the study objective, this design
Page 1165
www.rsisinternational.org
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue II, February 2026
provided a practical framework for examining the relationship between demand forecasting practices and
operational performance in public hospitals, thereby generating empirical evidence on how forecasting
capabilities are linked to service delivery and efficiency outcomes within the county’s healthcare system.
Area of study
Siaya County is one of Kenya’s 47 devolved units and is situated in the former Nyanza region in the western
part of the country. The county is administratively divided into six sub-counties: Ugenya, Ugunja, Gem, Bondo,
Rarieda, and Alego-Usonga. Siaya County was selected as the study site due to its documented public health
challenges. Reports from the Ministry of Health (2018) indicated that the county recorded among the highest
mortality rates in the country, with residents facing an elevated risk of premature deaths largely associated with
the high burden of HIV/AIDS and malaria. Furthermore, data from the World Health Organization (2017) ranked
Siaya among the counties with the highest HIV prevalence rates nationally, second only to Homa Bay County.
These epidemiological realities underscore the critical importance of ensuring consistent availability of essential
medicines and efficient supply chain systems within public hospitals. The high disease burden in the county
makes it an appropriate context for examining how demand forecasting practices are linked to operational
outcomes in public healthcare facilities.
Target Population of the Study
The study targeted a total population of 106 personnel drawn from key functional departments involved in supply
chain and operational management across six sub-county hospitals in Siaya County. These facilities included
Sigomere Sub-County Hospital, Bondo Sub-County Hospital, Malanya Sub-County Hospital, Yala Sub-County
Hospital, Madiany Sub-County Hospital, and Siaya County Referral Hospital. The target population comprised
staff from procurement, stores, pharmacy, and hospital administration, as these departments play a central role
in demand forecasting and supply chain coordination. Specifically, the population consisted of 34 administrators,
33 pharmacists, 18 procurement officers, and 21 storekeepers. The distribution of the total target population of
106 respondents is presented in Table 2.1.
Table 2.1: Target Population
Section
Ungenya
(Sigomere
SubCounty
Hospital)
Ugunja
(Malanya
SubCounty
Hospital)
Gem
(Yala
SubCounty
Hospital)
Bondo
(Bondo
SubCounty
Hospital)
AlegoUsonga
(Siaya
County
Referral
Hospital)
Total
1.
Administration
05
05
05
05
09
34
2.
Pharmacy
06
04
03
05
11
33
3.
Procurement
03
03
02
03
05
18
4.
Stores
03
03
03
03
06
21
Total
106
Source: Siaya County MoH, (2025)
Sample Size and Sampling Procedure
This section outlines the determination of the study’s sample size and describes the sampling procedures
employed. The details are presented in the subsections that follow:
Sample Size
The sample size is a representative of a large population (Bryman, 2012). Yamane, (1967) formula was used in
determining the sample size. The sample size in each stratum was obtained proportionately. In the field the
respondents were selected using random sampling.
Page 1166
www.rsisinternational.org
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue II, February 2026
According to Yamane, (1967): n=
2
1 Ne
N
…………………………………Eq.2.1
Where n = is the sample size
N = is the population e = is the error limit (0.05 on the basis of
95% confidence level)
Therefore, n = 106 / [1 + 106 (0.05)
2
]
n = 106/1.265
n = 84
Using a population of 106 staff members in public hospitals in Siaya County and considering an error limit of
5%, a sample size of 84 was used in the study. This sample size was representative enough and was spread in
each stratum proportionately as illustrated in Table 2.2.
Table 2.2: Sample Frame
Section
Population
(X)
Sample
Size
X/N x
84
Sigomere
Malanya
Yala
Bondo
Madiany
Siaya
County
Referral
Hospital)
1.
Administration
34
27
4
4
4
4
4
7
2.
Pharmacy
33
26
5
3
2
4
3
9
3.
Procurement
18
14
2
2
2
2
2
4
4.
Stores
21
17
2
2
2
3
3
5
Total
106
84
13
11
10
13
12
25
Source: Researcher’s own conceptualization, (2025)
Sampling Procedure
In view of the researcher’s inability to reach out to the entire population, and in order to gain the advantage of
an in-depth study and effective coverage, Yamane formula was used to establish the sample size from the study
population.
Stratified proportionate sampling was used to get sample size for each stratum. In the field, respondents were
selected using simple random sampling.
Data Collection Instruments
The study used structured questionnaires and interview guide in collecting primary data.
Questionnaires
The questions were based on a 5-point Likert scale. The questionnaire would capture information on the variables
and was divided into sections. Preliminary section captured general information of the respondents. The other
sections covered information on operational performance, and finally the last section covered information on
Demand Forecasting in Siaya County public hospitals.
Page 1167
www.rsisinternational.org
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue II, February 2026
Interview Guide
Interview with randomly selected respondents from sample size was used to compliment information derived
from the questionnaire items. This was an enhancement to gather information that may not have otherwise been
anticipated during the construction of the questionnaire.
Data Collection Procedure
Prior to data collection, the researcher obtained an introductory letter from Jaramogi Oginga Odinga University
of Science and Technology to facilitate the acquisition of a research permit from the National Commission for
Science, Technology and Innovation (NACOSTI). This authorization allowed access to the selected hospitals for
field data collection. Structured questionnaires were distributed to respondents using a drop-and-pick-later
approach at their respective workstations to minimize disruption of hospital operations. Follow-up was
conducted through telephone calls to remind participants of the agreed collection dates, after which the
completed questionnaires were retrieved. In addition, group interviews were organized with selected categories
of staff to complement the survey data. Upon collection, all questionnaires were reviewed to ensure completeness
and accuracy of responses before proceeding to data coding and analysis.
Pilot Testing
A pilot study was carried out at Kombewa Sub-County Hospital involving staff drawn from the four departments
targeted in the main study. Pilot studies are widely recommended to test feasibility and improve instrument
quality (van Teijlingen & Hundley, 2001). The purpose of the pilot test was to assess the clarity, relevance, and
reliability of the research instruments, as well as to estimate the time required to complete the questionnaire and
identify any structural or content-related weaknesses. According to Abuya (2018), between 10% and 30% of the
intended sample is adequate for pilot testing in survey research. In this study, 10% of the anticipated respondents
were selected for the pilot phase. Consequently, 10 staff members from Kombewa Sub-County Hospital were
randomly chosen to participate in the pre-test. Following the pilot exercise, the researcher examined response
patterns, evaluated participant feedback, and analyzed preliminary data to detect ambiguities or inconsistencies.
The insights obtained were then used to refine and improve the final data collection instruments before the main
study was undertaken.
Validity Test
Content validity was used to determine the validity index. Content validity measurement are used to emphasize
clarity on what CVI reflects and how it is reported (Polit & Beck, 2006). The questionnaires were given to the
two supervisors in the School of Business and Economics to evaluate and rate each item in relation to the
objectives as “not relevant” or “relevant” on a scale of the 1-4 such that; 1 = not relevant, 2 = somehow relevant,
3 = relevant and 4 = very relevant. Content validity index would then be determined from the supervisors’
agreement scale as K/N, where K being the number of items marked 3 or 4 and N the total number of items
assessed.
The rated finding was used to calculate content validity index (CVI) using the formula:
CVI =
K
N
……………………………………………………. Eq.2.2
Where: K = Total number of items in the questionnaire declared valid by both experts; and N = Total number of
items in the questionnaire. This was solved as follows:
CVI =
K
N
= = 0.8125
The computed instrument content validity index (CVI) was ε=0.8125 > ε=0.7. The computed CVI was
greater than the minimum acceptable index of 0.70 as recommended in the survey studies by Amin, (2005)
hence the instrument was valid for the study.
Page 1168
www.rsisinternational.org
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue II, February 2026
Reliability Test
Reliability of the instrument was checked through split-half reliability coefficient test. The items in the
questionnaire was divided into; odd items represented by “x” and even items represented by “y”. The scores
from both halves would then be correlated. Usually, the internal consistency of a measurement scale is assessed
by using Cronbach’s co-efficient alpha (Cronbach 1951) which was calculated using Flanagan Formula shown
in Eq. 2.3.
]1[2
2
2
2
2
1
t
t
R
……………………………………. Eq.2.3
Where: R
t
= Reliability Coefficient of the Test; δ
1
= Standard Deviation (S.D.) of Scores of 1
st
Half; δ
2
= Standard
Deviation (S.D.) of Scores of 2
nd
Half; and δ
t
= Standard Deviation (S.D.) of Scores of Whole Tests
For overall analysis on reliability using Cronbach’s alpha, the items analysed for this study were summed to
create the different scores, which formed a scale on which Cronbach's alpha was computed. The computed
Reliability Coefficient of the instrument was checked against the minimum acceptable index of 0.70 as
recommended in the survey studies by Nunnally and Bernstein (1994).
Reliabilities ranging from 0.5 to 0.60 are usually sufficient for exploratory studies (Nunnally & Bernstein, 1994),
while those in the range of 0.70 are acceptable and over 0.80 are good (Sekaran, 2003). The reliability obtained
was as summarized in Table 3.3a and Table 2.3b.
Table 2.3(a) Reliability Statistics
Cronbach's Alpha Based on
Un-Standardized Items
Cronbach's Alpha Based on Standardized Items
No. of Items
0.834
0.821
48
Source: Survey Data (2025)
Table 2.3(b) Reliability Statistics
Factor
No. of Items
Cronbach's Alpha
Based on UnStandardized Items
Cronbach's Alpha Based
on Standardized Items
Background
03
0.987
0.983
Demand Forecasting
08
0.863
0.854
Operational Performance
06
0.835
0.819
Overall
48
0.834
0.821
Source: Survey Data (2025)
The overall alpha for the Demand Forecasting items under investigation had a Cronbach’s alpha of 0.854
indicating good internal consistency, operational performance had a good Cronbach’s alpha coefficient of 0.819.
The minimum alpha for the items was 0.819 while the highest alpha was 0.983 both of which conformed to the
project by George and Mallery (2003) thus the items formed a scale that had excellent internal consistency
reliability.
Page 1169
www.rsisinternational.org
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue II, February 2026
Data Analysis
Upon completion of data collection, the responses were carefully edited, categorized, coded, and entered into a
computer for analysis. Statistical analysis was conducted using the Statistical Package for Social Sciences
(SPSS), Version 24, which is widely recognized for its robust data management capabilities and ability to perform
diverse statistical procedures suitable for both small and large datasets (Muijs, 2004). The study generated both
qualitative and quantitative data. Qualitative data obtained from interviews were analyzed using thematic and
content analysis techniques to identify recurring patterns and key insights. Quantitative data were analyzed using
descriptive and inferential statistical methods. Descriptive statistics included graphical and numerical summaries
such as frequencies, percentages, means, and standard deviations to describe the characteristics of the data.
Inferential analysis involved correlation and regression techniques to examine the relationship between demand
forecasting practices and operational outcomes. Correlation analysis was used to determine the strength and
direction of the linear association between variables, while regression analysis was applied to estimate the
predictive effect of demand forecasting on operational performance (Mutai, 2000). In addition, Analysis of
Variance (ANOVA) was performed to test the overall significance of the regression model. Individual regression
coefficients were further examined to assess the extent to which demand forecasting practices significantly
influenced operational performance in public hospitals. The study hypothesis was formulated to guide this
analytical process as follows:
H
O1
: Demand Forecasting has no statistically significant effect on operational performance of public hospitals
in Siaya County.
The following regression model to establish the relationship between the study variables guided the study:
Y= β
0
+ β
1
X
1
+e………………………………………Eq.2.4
Where: Y = Operational Performance; X
1
= Demand Forecasting Practice; e- Error Term; β
0
-represents the
Model Constant; and β
1 -
are Regression Coefficients.
The regression model assumed independent, identical and normally distributed random variables with a zero
mean and a constant variance at 5% significance level.
Diagnostic Tests for Inferential Statistics
Prior to inferential analysis, a series of diagnostic tests were conducted to verify that the data satisfied the
underlying assumptions of regression analysis. These assessments included tests for normality, multicollinearity,
homoscedasticity, and linearity. Normality of the data distribution was evaluated using the Kolmogorov–
Smirnov (K–S) test as part of exploratory data analysis. Numerical normality tests compare observed sample
scores with those expected under a normal distribution.
The K–S test is particularly appropriate for sample sizes greater than 50. A non-significant result (p > 0.05)
indicates that the data do not significantly deviate from normality. In this study, the K–S test yielded p = 0.55,
which exceeds the 0.05 threshold, confirming that the data were normally distributed. Multicollinearity among
independent variables was assessed using pairwise correlation analysis together with Tolerance and Variance
Inflation Factor (VIF) statistics. All correlation pvalues exceeded 0.05, indicating no significant intercorrelation
among the predictors. This result confirmed the absence of multicollinearity and supported the suitability of the
variables for regression modelling. Homoscedasticity was examined using the Breusch–Pagan approach based
on the observed R-squared statistic. The resulting p-value (0.3285) was greater than 0.05, leading to acceptance
of the null hypothesis of constant error variance. This indicates that the residuals were homoscedastic, a desirable
condition for reliable regression estimates. Linearity was assessed by plotting standardized residuals against
predicted values. The scatterplot showed no systematic curvature or bowed pattern, and the residuals were
symmetrically distributed around the horizontal axis with approximately constant spread. This pattern confirmed
that the relationship between the study variables satisfied the linearity assumption required for regression
analysis.
Page 1170
www.rsisinternational.org
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue II, February 2026
FINDINGS
Descriptive Statistics on Demand Forecasting
Descriptive analysis was done on the effect of demand forecasting practice on operational performance. The
results were summarized in table 3.1a
Table 3.1a: Descriptive Statistics for Demand Forecasting Practice and Operational Performance
Demand Forecasting and Operational Performance
(Operational Efficiency)
Response
N
%
Frequency
%
Frequency
Mean
Std.
Dev
Our constant usage of proper MRP System in our
inventory management have led to improved operational
efficiency through the pharmacy that is always well
80
75 (60)
25 (20)
2.754
0.354
stocked with required drugs
Our constant usage of proper EOQ Model in our
inventory management have led to improved operational
efficiency through the pharmacy that is always well
80
69 (55)
31 (25)
2.769
0.332
stocked with required drugs
The use of MRP System in our inventory management has
led to well stocked surgical and non-surgical inventory
required by medics hence improved
80
87 (70)
13 (10)
2.716
0.318
The use of EOQ Model in our inventory management has
led to well stocked surgical and non-surgical inventory
required by medics hence improved
80
84 (67)
16 (13)
2.675
0.322
AVERAGE
79
21
2.136
0.129
Demand Forecasting and Operational Performance
(Service Delivery)
N
(63) %
Frequency
(17) %
Frequency
Mean
Std.
Dev
Our constant usage of proper MRP System in our
inventory management have led to improved operational
efficiency through the pharmacy that is always well
80
77 (62)
23 (18)
2.625
0.401
stocked with drugs required hence improved service Our
constant usage of proper EOQ Model in our inventory
management have led to improved operational efficiency
through the pharmacy that is always well
80
81 (65)
19 (15)
2.763
0.315
stocked with drugs required hence improved service The
use of MRP System in our inventory management has led
to well stocked surgical and non-surgical inventory
required by medics hence improved service
80
85 (68)
15 (12)
2.469
0.116
The use of EOQ Model in our inventory management has
led to well stocked surgical and non-surgical inventory
required by medics hence improved service
80
83 (66)
17 (14)
2.845
0.347
AVERAGE
82 (66)
18 (12)
2.676
0.295
Source: Survey Data (2025)
The study sought to investigate the effect of Demand Forecasting practice on operational performance of public
hospitals in Siaya County. Table 3.1a shows that majority of Siaya County public hospitals believe that Demand
Page 1171
www.rsisinternational.org
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue II, February 2026
Forecasting would have an effect on operational performance of public hospitals in Siaya County, with a mean
response of 2.136 (for operational efficiency) within the range of 2.675 ≤ µ ≤ 2.769 at 79% (S.D=.129) and a
mean response of 2.676 (for service delivery) within the range of 2.469 ≤ µ ≤ 2.845 at 82% (S.D=.295). This
finding is consistent with the findings of Kalchschmidt (2014), Louis (2015), Ngai-Hang et al (2016) and Oballa
et al (2015). Kalchschmidst (2014) established that demand forecasting is considered crucial process for
effectively guiding several activities within a manufacturing industry. Ngai-Hang et al (2016) found out that
manual guesswork in forecasting demand is directly liked to inventory stock outs, the study recommended
scientific demand forecasting practices to be adopted to experience improved performances. Oballa et al (2015)
established that having accurate future demand would lead to a positive influence on the organizational
performances. Louis (2015) recommended proper demand forecasting with efficient coordination within the
supply chain.
Descriptive Statistics on Operational Performance
Descriptive analysis was done on operational performance. The results were summarized in table 3.1b
Table 3.1b: Descriptive Statistics for Operational Performance
Operational Performance (Operational
Efficiency)
N
%
Frequency
(Agree)
%
Frequency
(Disagree)
Mean
Std.
Dev
The hospital always handles a large number of
out- patient’s cases on daily basis
80
55(44)
45(36)
2.933
1.216
It takes the shortest time possible for patients to
go through the treatment process (for out-patient)
80
58(46)
42(34)
2.628
1.137
For the last two years the hospital has recorded
reduction in mortality rates
80
56(57)
44(43)
2.917
1.313
In-patients cases normally stay in the hospitals for
a shorter period before being discharged
80
51(41)
49(39)
3.007
1.144
AVERAGE
55(44)
21(17)
2.871
1.203
Performance in Terms of Service Delivery
N
%
Frequency
(Agree)
%
Frequency
(Disagree)
Mean
Std.
Dev
The suggestion box is easily accessible to patients
80
60(48)
40(32)
3.726
1.321
On average, patients are always satisfied with the
hospital services
80
51(41)
49(39)
4.238
1.421
Action is always taken on feedbacks from the
suggestion box
80
76(61)
24(19)
2.416
1.232
On average, patients do get the prescribed drugs
in the hospital pharmacy
80
87(70)
13(10)
2.118
1.279
AVERAGE
69(55)
31(25)
3.125
1.313
Source: Survey Data (2025)
Table 3.1b shows that majority of Siaya County public hospitals believe that operational performance of public
hospitals in Siaya County is good. Specifically, operational efficiency had a mean response of 2.871 within the
range of 2.628 ≤ µ ≤ 3.007 at 55% (S.D=1.203). This implies that 55% of the respondents in Siaya County public
hospitals do agree that operational efficiency was considerably good. In addition, service delivery was also
considerably good at a mean response of 2.522 within the range of 2.118 ≤ µ ≤ 2.726 at 69% (S.D=1.313).
Page 1172
www.rsisinternational.org
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue II, February 2026
Inferential Statistics
Hypothesis stated that there is no significant statistical effect of Demand Forecasting practice on operational
performance of public hospitals in Siaya County. The effect of Demand Forecasting on operational performance
of public hospitals in Siaya County was investigated through linear regression analysis using the model in
equation 3.0:
1110
XP
……………………………………………………………Eq. (3.0)
where P is operational performance, while β
0
is the intercept (a constant), β
1
, is the slope associated to the
independent variables X
1
, and ε is the error term which is assumed to be independent, identical normally
distributed random variable with a zero mean and a constant variance. The results were as captured in
Table 3.2, Table 3.3, and Table 3.4.
Table 3.2 Model Summary
Model
R
R Square
Adjusted R Square
Std. Error of the Estimate
1
.837
.701
.678
.021
Source: Survey Data (2019)
The Model Summary table 3.2 gives the R square (0.701) and Adjusted R square (0.678). Thus, in this model,
demand forecasting is predicting 70.1% of the variance in operational performance of public hospitals in Siaya
County.
This leaves 29.9% of the variation in operational performance of public hospitals in Siaya County being
explained by the error-term or other variables other than demand forecasting. This finding also indicated the
model’s goodness of fit as exemplified by the coefficient of determination value of (R2 value) of 0.701 adjusted
to of 0.678.
The standard error of the estimate, of 0.021 being a measure of standard deviation around the fitted line suggests
that about 95% of the prediction error in demand forecasting - operational performance model of public hospitals
in Siaya County is less than ±1.96 (0.046) = 0.041.
Table 3.3
ANOVA
Model
df
F-Change
Sig. F-Change
Durbin - Watson
1
Regression
1
9.195
0.001
1.602
Residual
79
Total
80
Source: Survey Data (2025)
The ANOVA table 3.3 shows that the computed F statistic was 9.195, with an observed significance level (pvalue)
of 0.001 which was also less than p<0. 05. This shows that the significance can be extended to 0.01, or 99.99%
confidence interval.
The independence of residuals in this model was analysed using Durbin-Watson statistic. Considering a Durbin-
Watson statistic of 1.602, it was deduced that there was no serial correlation of the residuals as the values were
within the accepted threshold of between 1.5 to 2.5 as was recommended by Hayes, (2013).
Page 1173
www.rsisinternational.org
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue II, February 2026
Table 3.4: Regression Coefficients of Demand Forecasting and Operational Performance
Model
Unstandardized Coefficients
Stand. Coef.
t
Sig.
B
Std. Er.
Beta
1(Constant)
Demand Forecasting
2.382
.820
2.905
.000
.876
.191
.685
4.578
.000
R
0.837
R-squared
0.701
Adjusted R-squared
0.678
F-statistics
9.195
Prob(F-statistics)
0.001
Source: Survey Data (2025)
Considering the hypothesis that there is no statistically significant effect of demand forecasting practices on the
operational performance of public hospitals in Siaya County, the results of the regression analysis in, Table 3.4,
revealed a strong and statistically significant positive relationship between demand forecasting and operational
performance (β = 0.876, p < 0.05). This indicates that improvements in demand forecasting practices are
associated with corresponding improvements in hospital operational outcomes. Specifically, the regression
coefficient suggests that a one-unit increase in demand forecasting capability is associated with a 0.876-unit
improvement in operational performance among public hospitals in Siaya County. Consequently, the null
hypothesis was rejected and the alternative hypothesis, which posits that demand forecasting practices
significantly influence operational performance, was accepted.
The findings of this study provide empirical support for the theoretical propositions of the Resource-Based View
(RBV) and Network Perspective Theory in explaining operational outcomes within healthcare supply chains.
From the RBV perspective, demand forecasting capability can be conceptualized as a strategic organizational
resource that enhances institutional performance. RBV suggests that organizations achieve superior outcomes
when they effectively deploy valuable, rare, inimitable, and well-organized capabilities (Barney, 1991). In the
context of public hospitals, forecasting systems, inventory planning tools, data management practices, and skilled
supply chain personnel constitute critical internal capabilities that determine how effectively healthcare facilities
plan procurement and manage pharmaceutical inventories. The strong explanatory power of the regression model
(R² = 0.701) further reinforces the argument that forecasting capability represents a critical operational resource
in healthcare supply chain systems.
The findings demonstrate that hospitals that systematically apply structured forecasting techniques, such as
demand planning models, inventory control systems, and data-driven procurement processes, are better
positioned to maintain optimal pharmaceutical stock levels and minimize service disruptions associated with
medicine shortages or excess inventory. These findings align with earlier studies that emphasize the strategic
importance of forecasting for operational performance. For example, Kalchschmidt (2014) established that
forecasting is a crucial mechanism for coordinating supply chain activities and guiding operational planning.
Similarly, Oballa et al. (2015) demonstrated that accurate demand forecasting contributes significantly to
improved performance in healthcare institutions. Other studies have also shown that effective forecasting reduces
stock-out costs, enhances inventory optimization, and improves service delivery outcomes (Ngai-Hang et al.,
2016; Kisaka, 2016; Louis, 2015; Haruna, 2019). It is therefore evident that structured forecasting improves
operational planning and inventory outcomes (Chopra & Meindl, 2021; Fildes et al., 2009). Additionally,
evidence from health supply systems links weak inventory management to stock-outs and service disruptions
(Leung et al., 2016; Toroitich et al., 2022).
Page 1174
www.rsisinternational.org
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue II, February 2026
Beyond traditional forecasting methods such as Material Requirements Planning (MRP) and Economic Order
Quantity (EOQ), recent developments in forecasting highlight the role of machine learning and advanced
analytics in improving predictive performance (Sezer et al., 2020). In healthcare supply chains, analytics-driven
decision support has been increasingly reviewed as a pathway to improve supply responsiveness and operational
efficiency (Subramanian, 2021). Contemporary research suggests that AI-driven forecasting systems integrate
historical consumption data, epidemiological trends, and real-time demand signals to generate more accurate
predictions of healthcare supply requirements. These technologies enable healthcare organizations to optimize
inventory management, improve demand forecasting, and streamline procurement processes, thereby enhancing
operational efficiency and service delivery outcomes.
Similarly, predictive analytics tools have been shown to transform healthcare supply chains by improving
inventory management and enabling more responsive procurement systems. By leveraging machine learning
algorithms and real-time data analytics, healthcare supply chains can significantly enhance forecasting accuracy
and operational efficiency. In pharmaceutical supply chains, AI-driven forecasting models can outperform
traditional forecasting techniques by identifying complex demand patterns and anticipating potential supply
disruptions. These advancements suggest that integrating digital forecasting technologies into hospital supply
chain systems could significantly strengthen pharmaceutical availability and operational resilience.
The relevance of these findings becomes even more significant when viewed through the lens of Network
Perspective Theory, which emphasizes the role of collaborative relationships and information exchange among
supply chain actors in influencing organizational performance. Healthcare supply chains are inherently
networked systems involving multiple stakeholders including hospitals, pharmaceutical suppliers, procurement
agencies, regulatory institutions, and central distribution bodies. Effective demand forecasting therefore depends
not only on internal organizational capabilities but also on the quality of coordination and communication across
the healthcare supply network.
In the Kenyan context, hospitals depend heavily on the Kenya Medical Supplies Authority (KEMSA) for
procurement and distribution of essential medicines. Consequently, forecasting accuracy within hospitals is
closely linked to the effectiveness of information sharing and coordination between hospitals and central supply
agencies. Weak communication within this network can lead to procurement delays, inaccurate demand
projections, and frequent stock-outs of essential medicines.
Qualitative evidence obtained from respondents in this study reinforces this interpretation. One respondent
observed that although hospitals attempt to forecast pharmaceutical demand, unpredictable disease outbreaks
such as epidemics and pandemics often distort historical demand patterns, resulting in cases of both overstocking
and understocking of medicines. In such situations, hospitals may procure quantities that either exceed or fall
below actual demand levels, thereby generating inefficiencies in inventory management and negatively affecting
service delivery outcomes. Another respondent emphasized that improving demand forecasting systems could
significantly reduce cases where patients leave hospitals without receiving prescribed medicines, highlighting
the direct relationship between forecasting capability and healthcare service accessibility.
Kenya’s pursuit of Universal Health Coverage has highlighted persistent operational constraints, including
medicine availability and facility readiness (Walukana et al., 2021). Stock-outs and affordability barriers can
undermine effective access even where service utilization improves (Toroitich et al., 2022). Universal health
coverage aims to ensure that all individuals have access to quality healthcare services without financial hardship.
Kenya has prioritized UHC as a major health sector reform through national policy initiatives and pilot programs
implemented since 2018. However, several studies have noted that persistent supply chain challenges,
particularly medicine shortages and delayed procurement processes, continue to undermine the effectiveness of
UHC implementation. For instance, evidence from Kenya’s UHC pilot phase shows that although access to
healthcare services improved, facilities continued to experience challenges related to inadequate medical
supplies and inconsistent pharmaceutical availability.
In this context, strengthening demand forecasting capabilities within hospital supply chains becomes a critical
policy priority for achieving sustainable universal health coverage. Accurate forecasting systems can help ensure
that essential medicines are available in the right quantities, at the right time, and in the right locations, thereby
Page 1175
www.rsisinternational.org
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue II, February 2026
supporting continuous service delivery within public health facilities. Furthermore, integrating advanced
forecasting technologies such as predictive analytics and AI-based demand modelling could enable healthcare
systems to better anticipate disease outbreaks, respond to fluctuations in patient demand, and enhance supply
chain resilience.
Overall, the findings of this study suggest that improving demand forecasting systems, through investments in
digital forecasting technologies, enhanced data integration, and stronger coordination across healthcare supply
networks, can significantly improve operational performance in public hospitals. By strengthening both internal
forecasting capabilities (as emphasized by RBV) and external supply chain collaboration mechanisms (as
emphasized by Network Perspective Theory), healthcare institutions can improve drug availability, reduce
service disruptions, and ultimately enhance the effectiveness of healthcare delivery within Kenya’s evolving
universal health coverage framework.
SUMMARY OF FINDINGS
The objective sought to establish the effect of Demand Forecasting on Operational Performances of public
hospitals in Siaya County. 79% (Mean 2.136: SD=.129) of the public hospital workers believe that demand
forecasting influences the level of operational efficiency in public hospitals. 82% (Mean 2.676: SD=.295) of the
public hospital workers believe that demand forecasting influences the level of service delivery in public
hospitals. The Hypothesis stated that there is no significant statistical effect of demand forecasting practice on
operational performance of public hospitals in Siaya County. This was, however, rejected based on the findings
which showed that demand forecasting had a statistically significant effect on performance of public hospitals
in Siaya County with a coefficient of β=.876. This implies that when keeping the effects of other factors constant,
a unit increase in demand forecasting would increase operational performances of public hospitals in Siaya
County by 0.876 units.
CONCLUSION
The study sought to establish the effect of Demand Forecasting on Operational Performances of public hospitals
in Siaya County. The study finding indicated that Demand Forecasting has statistically significant effect on
operational performances of public hospitals in Siaya County. From the findings obtained herein, it was
concluded that the efforts a public hospital put in Demand Forecasting as an inventory management practice
would eventually become critical in realizing an improved operational performance in terms of service delivery
for the patients. Further to this, it was also evident that Demand Forecasting does not necessarily determine
operational performance independently but rather in addition to other inventory management practices.
RECOMMENDATION
Objective one sought to establish the effect of Demand Forecasting on Operational performances of Public
hospitals in Siaya County. The study thus recommends that accurate demand forecasts should always be done to
avoid service delivery disruptions.
Implications for Theory and Practice
Implications for Theory
This study contributes to the growing body of literature on healthcare supply chain management by providing
empirical evidence on the role of demand forecasting capabilities in shaping operational outcomes within public
hospitals. By integrating the Resource-Based View (RBV) and Network Perspective Theory, the study extends
theoretical understanding of how internal organizational capabilities and external supply chain relationships
jointly influence healthcare system performance. From the perspective of the Resource-Based View, the findings
reinforce the argument that organizational performance improvements are strongly linked to the effective
deployment of strategic capabilities rather than merely the possession of resources. In this study, demand
forecasting capability emerges as a critical strategic resource within hospital supply chain systems. Forecasting
tools, data management systems, and skilled personnel collectively constitute an operational capability that
Page 1176
www.rsisinternational.org
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue II, February 2026
enables hospitals to anticipate pharmaceutical demand and manage inventory more efficiently. The statistically
significant relationship between demand forecasting and operational performance supports RBV’s proposition
that well-developed organizational capabilities can generate superior performance outcomes. The findings also
expand the application of RBV within the healthcare supply chain context. While RBV has traditionally been
applied within manufacturing and private sector organizations, this study demonstrates its relevance within
public healthcare institutions where operational capabilities, such as forecasting systems and inventory
management practices, play a critical role in service delivery outcomes. In addition, the study contributes to
theory by highlighting the complementary role of Network Perspective Theory in explaining healthcare supply
chain performance. Healthcare supply chains operate within complex institutional networks involving hospitals,
procurement agencies, pharmaceutical suppliers, regulatory bodies, and logistics providers. Effective forecasting
therefore depends not only on internal organizational competencies but also on the strength of relationships and
information exchange across the supply network. The findings, therefore, demonstrate that forecasting
effectiveness is partly influenced by coordination between hospitals and supply chain actors such as the Kenya
Medical Supplies Authority (KEMSA). Weak coordination or delayed information sharing within this network
can reduce forecasting accuracy and contribute to medicine shortages. By illustrating how forecasting capability
operates both as an internal resource and a network-dependent operational process, this study contributes to the
theoretical integration of RBV and Network Perspective Theory within healthcare supply chain research.
Furthermore, the study highlights the growing importance of digital supply chain capabilities, including
predictive analytics, artificial intelligence, and integrated health logistics information systems, in improving
forecasting accuracy. Incorporating such technological capabilities into theoretical frameworks may provide new
directions for future research on healthcare supply chain resilience and operational efficiency.
Implications for Practice
The findings of this study also have significant practical implications for healthcare managers, supply chain
practitioners, and policy makers responsible for improving healthcare service delivery within public health
systems. First, the results emphasize the importance of strengthening demand forecasting systems within public
hospitals. Hospital managers should prioritize the adoption of structured forecasting approaches supported by
reliable consumption data, inventory monitoring systems, and integrated supply chain planning processes.
Improving forecasting accuracy can help hospitals maintain optimal stock levels, minimize medicine shortages,
and enhance service delivery efficiency.
Second, the study highlights the need for capacity building among healthcare supply chain personnel. Forecasting
systems are only effective when supported by skilled staff capable of interpreting demand data and applying
forecasting models in procurement planning. Continuous training programs for pharmacists, procurement
officers, and supply chain managers can significantly enhance forecasting capability within healthcare
institutions.
Third, the findings underscore the importance of digital transformation within healthcare supply chains. Many
public hospitals still rely on manual forecasting methods that are unable to capture complex demand patterns
driven by epidemiological trends and seasonal disease outbreaks. Integrating digital forecasting tools such as
predictive analytics platforms, Electronic Logistics Management Information Systems (e-LMIS), and artificial
intelligence-based demand modelling can significantly improve forecasting accuracy and supply chain
responsiveness.
Fourth, the study highlights the importance of strengthening coordination within the healthcare supply network.
Effective forecasting requires continuous communication and information sharing between hospitals,
procurement agencies, suppliers, and distribution bodies. Strengthening coordination mechanisms between
public hospitals and KEMSA could significantly reduce procurement delays and improve the availability of
essential medicines.
Finally, the findings have important implications for health policy and the implementation of Universal Health
Coverage (UHC) in Kenya. One of the key objectives of UHC is to ensure equitable access to essential healthcare
services and medicines. However, persistent medicine stock-outs continue to undermine service delivery in many
public hospitals. Strengthening forecasting systems within hospital supply chains can therefore play a critical
Page 1177
www.rsisinternational.org
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue II, February 2026
role in supporting the successful implementation of UHC by ensuring the continuous availability of essential
medicines and medical supplies.
Policy makers should therefore consider investing in national forecasting systems, integrated health supply chain
data platforms, and digital logistics infrastructure that enable real-time monitoring of medicine demand and
supply. Such investments would enhance healthcare system resilience and improve the efficiency of
pharmaceutical supply chains across the country.
Implications for Future Research
This study also opens several avenues for future research. First, future studies could examine the role of advanced
forecasting technologies, such as machine learning and artificial intelligence, in improving pharmaceutical
demand forecasting within healthcare supply chains. Second, longitudinal studies could provide deeper insights
into how forecasting capability influences operational performance over time. Third, comparative studies across
different counties or national healthcare systems could help identify contextual factors that influence forecasting
effectiveness within public health supply chains.
REFERENCES
1. Abdel-Maksoud, A., Dugdale, D., & Luther, R. (2008). The role of performance measurement in
continuous improvement. International Journal of Operations & Production Management, 28(7), 667– 692.
https://doi.org/10.1108/01443570810881858.
2. Abuya J. O. (2018). Cash Flow, Supply Chain Performance and Lead Time of Road Construction Projects
in Kenya. African Journal of Business and Industry, 2(4),
292-312. https://www.researchgate.net/search.Search.html?type=publication&query.
3. Asubonteng, P., McCleary, K. J., & Swan, J. E. (1996). SERVQUAL revisited: A critical review of service
quality. Journal of Services Marketing, 10(6), 62–81. https://doi.org/10.1108/08876049610148602 4.
Barney, J. B. (1991). Firm resources and sustained competitive advantage. Journal of Management, 17(1),
99–120. https://doi.org/10.1177/014920639101700108.
5. Borgatti, S. P., & Li, X. (2009). On social network analysis in a supply chain context. Journal of Supply
Chain Management, 45(2), 5–22. https://doi.org/10.1111/j.1745-493X.2009.03166.x
6. Bosua, R., & Scheepers, R. (2007). The role of social networks in knowledge sharing: A study of a
knowledge-intensive organisation. Australasian Journal of Information Systems, 14(2), 17–34.
https://doi.org/10.3127/ajis.v14i2.477
7. Bryman, A. (2012). Social research methods (4th ed.). Oxford University Press.
8. Chopra, S., & Meindl, P. (2021). Supply chain management: Strategy, planning, and operation (7
th
ed.).
Pearson.
9. Cooper, D. R., & Schindler, P. S. (2014). Business research methods (12th ed.). McGraw-Hill Education.
10. Cronbach, L. J. (1951). Coefficient alpha and the internal structure of tests. Psychometrika, 16(3), 297–
334. https://doi.org/10.1007/BF02310555.
11. de Vries, J., & Huijsman, R. (2011). Supply chain management in health services: An overview. Supply
Chain Management: An International Journal, 16(3), 159–165.
https://doi.org/10.1108/13598541111127146.
12. Fildes, R., Goodwin, P., Lawrence, M., & Nikolopoulos, K. (2009). Effective forecasting and judgmental
adjustments: An empirical evaluation and strategies for improvement in supply-chain planning.
International Journal of Forecasting, 25(1), 3–23. https://doi.org/10.1016/j.ijforecast.2008.11.010.
13. Flores, B. E., & Whybark, D. C. (1986). Multiple criteria ABC analysis. International Journal of Operations
& Production Management, 6(3), 38–46. https://doi.org/10.1108/eb054761.
14. Gefen, D. (2002). Customer loyalty in e-commerce. Journal of the Association for Information Systems,
3(1), 27–51. https://doi.org/10.17705/1jais.00022.
15. George, D., & Mallery, P. (2003). SPSS for Windows step by step: A simple guide and reference (4
th
ed.).
Allyn & Bacon.
16. Hayes, A. F. (2013). Introduction to mediation, moderation, and conditional process analysis: A regression-
based approach. The Guilford Press.
Page 1178
www.rsisinternational.org
INTERNATIONAL JOURNAL OF LATEST TECHNOLOGY IN ENGINEERING,
MANAGEMENT & APPLIED SCIENCE (IJLTEMAS)
ISSN 2278-2540 | DOI: 10.51583/IJLTEMAS | Volume XV, Issue II, February 2026
17. Hyndman, R. J., & Athanasopoulos, G. (2021). Forecasting: Principles and practice (3rded.). OTexts.
https://otexts.com/fpp3/
18. Kaplan, R. S., & Norton, D. P. (1992). The balanced scorecard—Measures that drive performance. Harvard
Business Review, 70(1), 71–79.
19. Kenya National Bureau of Statistics. (2021). Kenya poverty report 2021. Kenya National Bureau of
Statistics. https://www.knbs.or.ke/?wpdmpro=kenya-poverty-report-2021.
20. Leung, N. H. L., Chen, A., Yadav, P., Gallien, J., Durand, L., & Kearns, P. (2016). The impact of inventory
management on stock-outs of essential drugs in sub-Saharan Africa: Secondary analysis of a field
experiment in Zambia. PLOS ONE, 11(5), e0156026. https://doi.org/10.1371/journal.pone.0156026.
21. Lysons, K., & Farrington, B. (2016). Procurement and supply chain management (9th ed.). Pearson.
22. Ministry of Health (Kenya). (2014). Kenya health policy 2014–2030. Ministry of Health.
https://www.health.go.ke/wp-content/uploads/2015/09/Kenya-Health-Policy.pdf.
23. Muijs, D. (2004). Doing quantitative research in education with SPSS. SAGE Publications.
24. Neely, A. (2005). The evolution of performance measurement research: Developments in the last decade
and a research agenda for the next. International Journal of Operations & Production Management, 25(12),
1264–1277. https://doi.org/10.1108/01443570510633648.
25. Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric theory (3rd ed.). McGraw-Hill.
26. Parasuraman, A., Zeithaml, V. A., & Berry, L. L. (1988). SERVQUAL: A multiple-item scale for measuring
consumer perceptions of service quality. Journal of Retailing, 64(1), 12–40.
27. Penrose, E. T. (1959). The theory of the growth of the firm. Oxford University Press.
28. Polit, D. F., & Beck, C. T. (2006). The content validity index: Are you sure you know what’s being
reported? Critique and recommendations. Research in Nursing & Health, 29(5), 489–497.
https://doi.org/10.1002/nur.20147.
29. Ray, G., Barney, J. B., & Muhanna, W. A. (2004). Capabilities, business processes, and competitive
advantage: Choosing the dependent variable in empirical tests of the resource-based view. Strategic
Management Journal, 25(1), 23–37. https://doi.org/10.1002/smj.366.
30. Sekaran, U. (2003). Research methods for business: A skill-building approach (4th ed.). John Wiley &
Sons.
31. Sezer, O. B., Gudelek, M. U., & Ozbayoglu, A. M. (2020). Financial time series forecasting with deep
learning: A systematic literature review. Applied Soft Computing, 90, 106181.
https://doi.org/10.1016/j.asoc.2020.106181.
32. Stevenson, W. J. (2009). Operations management (10th ed.). McGraw-Hill/Irwin.
33. Stevenson, W. J. (2010). Operations management (11th ed.). McGraw-Hill/Irwin.
34. Subramanian, N. (2021). A systematic review on the integration of big data analytics in healthcare supply
chain management. Logistics, 5(4), 88. https://doi.org/10.3390/logistics5040088.
35. Toroitich, A. M., Chui, M. M., & Kelly, M. (2022). Patients’ access to medicines: A critical review of the
healthcare system in Kenya. Risk Management and Healthcare Policy, 15, 1937–1952.
https://doi.org/10.2147/RMHP.S373644.
36. van Teijlingen, E., & Hundley, V. (2001). The importance of pilot studies. Social Research Update, 35, 1–
4. University of Surrey. https://sru.soc.surrey.ac.uk/SRU35.html.
37. Wasserman, S., & Faust, K. (1994). Social network analysis: Methods and applications. Cambridge
University Press. https://doi.org/10.1017/CBO9780511815478.
38. Walukana, S., Lagat, A. K., Maina, I. W., & Tsofa, B. (2021). Examining the implementation experience
of the universal health coverage pilot in Kenya. BMJ Global Health, 6(9), e006309.
https://doi.org/10.1136/bmjgh-2021-006309.
39. Wernerfelt, B. (1984). A resource-based view of the firm. Strategic Management Journal, 5(2), 171–180.
https://doi.org/10.1002/smj.4250050207.
40. Yamane, T. (1967). Statistics: An introductory analysis (2nd ed.). Harper & Row.