The length of stay in hospital is broadly accepted as a measure of the burden of disease, of the efficiency and use of health services. While the duration of hospital stay is needed to stabilize and treat the patient, an unnecessarily prolonged hospital stay can lead to hospital acquired infections, physical deconditioning, medication related adverse effects and psychological distress, as well as increased hospital costs and a loss of hospital bed availability [1]. Patients who are admitted to the internal medicine units are especially at risk, as they usually have more than one chronic disease, rapid deterioration and complex diagnostic and therapeutic requirements. The length of hospitalization could thus be a reflection of the severity of the presenting illness or problems with hospitalisation and discharge planning [2, 3].

Prolonged hospitalization is a condition that varies across healthcare settings. Some studies set a specific criterion (e.g., > 7, 14 or 21 days) while others use the upper quartile or percentile of the distribution of observed length of stay. Although several different factors have been cited in the literature, older age, multimorbidity, functional impairment, prior hospitalisation, severe infection, renal dysfunction and malnutrition are all found to be significant factors in previous investigations which have linked to longer hospital stays [4]. Other indicators of disease burden, such as the Charlson Comorbidity Index, have also been correlated with prolonged hospital stay and intensive care use. None of the clinical or demographic variables, however, is sufficient to account for the length of hospital admission since length of stay is influenced by the interaction of all these variables [5].

There are some laboratory and clinical parameters that can help to identify patients who are likely to become long-stay patients. The complications of acute kidney injury include fluid imbalance, electrolyte disturbance, uraemic complications and renal replacement therapy, which can cause delay in recovery. Likewise, sepsis can result in haemodynamic instability, organ dysfunction and a prolonged need for antimicrobial and supportive treatment. Hypoalbuminaemia is linked to ill health, malnutrition, clinical outcome, and inflammation, and may indicate a decreased physiological reserve. Also, complications after admission, especially nosocomial infections, delirium, pressure ulcers, and a thrombotic event, can prolong the hospital stay, and may also contribute to mortality [6, 7].

Although several predictors of prolonged hospitalization have been reported internationally, their relative importance may vary according to patient characteristics, disease patterns, hospital facilities and local discharge practices. Evidence from internal medicine units in resource-limited settings remains comparatively limited. Identifying locally relevant predictors may help clinicians recognize high-risk patients soon after admission and initiate early multidisciplinary management, nutritional support, complication prevention and discharge planning. Therefore, this study was conducted to determine the frequency of prolonged hospitalization and evaluate the demographic, clinical, laboratory and hospital-related factors associated with extended hospital stay among patients admitted to internal medicine units.

This prospective cohort study was conducted in the internal medicine units of a Tertiary Care hospital of Islamabad from August 2025 to February 2026. The study aimed to identify demographic, clinical, laboratory and hospital-related predictors of prolonged hospitalization among adult patients admitted under internal medicine. All eligible patients were enrolled at the time of admission and followed prospectively until discharge, transfer, death or leaving the hospital against medical advice. Prolonged hospitalization was operationally defined as a hospital stay of more than 7 days, while a stay of 7 days or less was categorized as a non-prolonged hospitalization. The duration of hospitalization was calculated from the date of admission to the date of discharge or other final hospital outcome. Through consecutive non-probability sampling, 93 patients were included. The sample size was calculated using the formula for a single population proportion with an anticipated proportion of prolonged hospitalization among medical inpatients, a 95% confidence level and an acceptable margin of error. Consecutive sampling of all patients who fulfilled the eligibility criteria was carried out throughout the study period until the sample size was reached. Adult patients (age 18 years or older) admitted to the internal medicine units and anticipated to be admitted for a minimum of 24 hours were included. Patients were excluded from the study who were admitted only for observation, elective short procedures and/or diagnostic testing. Those who were transferred from another hospital after several days of admission, those patients who did not have baseline information, patients who declined participation, patients discharged within 24 hours were also excluded. A structured pre tested data collection form was used by the research team in collecting the data. Data at admission included age, gender, residence, smoking status, history of admission, referral status, length of symptoms and previous hospitalization. The primary diagnosis was classified into respiratory, cardiovascular, infectious, renal, endocrine or metabolic, gastrointestinal or hepatic, neurological and other medical disorders. Documentation of diabetes mellitus, hypertension, ischaemic heart disease, heart failure, chronic kidney disease, chronic liver disease, chronic obstructive pulmonary disease, previous stroke, malignancy and other chronic conditions was obtained from the medical record and clinical assessment. All comorbidities were summed and multimorbidity was defined as having two or more chronic conditions. The Charlson Comorbidity Index was computed where data were available to provide a measure of overall burden of comorbidities. Upon admission the clinical status of each patient was evaluated based on their blood pressure, pulse, respiration, temperature, oxygen saturation and consciousness level. The presence of sepsis, respiratory distress, acute kidney injury, hypotension, altered mental status and the need for oxygen therapy was documented. Baseline laboratory data available and clinically indicated included haemoglobin, total leukocyte counts, platelet counts, serum sodium, potassium, urea, creatinine, albumin, blood glucose, liver function tests and C-reactive protein. Information recorded during hospitalization included the use of intravenous antibiotics, number of medications prescribed, the use of oxygen, blood transfusion, dialysis, mechanical ventilation, and specialist consultation or transfer to a high dependency or intensive care unit. In-hospital complications, including HAI, delirium, electrolyte imbalance, pressure wounds, VTE events, AKI events, adverse drug reactions, and other complications were assessed daily. The final outcomes were length of stay, clinical condition at discharge, transfer to another facility, discharge against medical advice and in-hospital mortality. The data were inputted and analysed using IBM SPSS Statistics version 26. All continuous variables were reported as mean ± SD if they were normally distributed and median (IQR) if they were not. Values for categorical variables were expressed as frequencies and percentages. The independent-samples t-test or Mann–Whitney U test was used for continuous variables and the chi-square test or Fisher's exact test was used for categorical variables between patients with hospital stays for 7 days or less, and those with hospital stays of more than 7 days. Those variables with a p value < 0.20 in the univariate analysis and clinically important variables were put into the multivariable binary logistic regression model. Odds ratios with 95% confidence intervals were adjusted to determine independent predictors of extended hospitalization. A Hosmer-Lemeshow goodness of fit test was used to assess model fitness and multicollinearity was checked prior to regression analysis. A p value < 0.05 was considered statistically significant.

A total of 93 patients admitted to the internal medicine units were included in the study. The mean age of the participants was 56.8 ± 15.2 years, (range 21-84 years). The majority of patients were 60 years or older (43.0%) and 35.5% were between 40-59 years of age. The study comprised males (52.7%) and females (47.3%) with 58.1% from rural areas. The median length of hospital stay was 6 days (IQ range 4–9 days). Thirty-four patients (36.6%) experienced prolonged hospitalization (operational definition: hospitalization > 7 days). The other 59 patients (63.4%) were discharged after 7 days. Patients with prolonged hospitalization had a mean length of stay of 11.4 ± 3.2 days while those without prolonged hospitalization had a mean length of stay of 5.1 ± 1.6 days.

Table 1. Baseline demographic characteristics of the study participants (n = 93)

The most common comorbid condition was hypertension (57.0%), diabetes mellitus (43.0%), and chronic kidney disease (19.4%). 43.0% (40 patients) had two or more coexisting medical conditions. The top two primary reasons for admission were respiratory diseases and cardiovascular diseases, before infectious diseases. Some of the most common individual diagnoses were pneumonia, acute exacerbation of chronic obstructive pulmonary disease, heart failure, sepsis and diabetic complications.

Table 2. Clinical diagnoses and comorbid conditions among the participants

Older patients were more likely to have prolonged hospitalizations (62.9 ± 13.2 years vs 53.3 ± 15.3 years; p = 0.003). Patients older than 60 years, previously admitted, with multiple comorbidities, sepsis, acute kidney injury and need for oxygen therapy were more likely to be hospitalized for a prolonged period of time. No significant correlation was found between sex and prolonged hospitalization, nor was there any correlation between the residential status and prolonged hospitalization, nor was there any correlation between smoking history and prolonged hospitalization. Patients' Charlson Comorbidity Index was significantly higher in those with extended hospitalization than in those with brief hospitalization (4.1 ± 1.7 vs 2.5 ± 1.5, p < 0.001). Likewise, the mean haemoglobin, serum albumin and levels of C-reactive protein and serum creatinine were lower in patients with longer hospitalisation. The prolonged-hospitalization group also had higher rates of hyponatraemia and leucocytosis.

Table 3. Comparison of patients according to duration of hospitalization

In-hospital complications occurred in 21.5% of the patients during their hospital stay. The most frequent complications were hospital-acquired infection, electrolyte disturbances, delirium and pressure injuries. Prolonged hospitalization (more than 7 days) was associated with a significantly higher proportion of complications than was hospitalization for 7 days or less (44.1% vs. 8.5%, p < 0.001). Twelve patients needed to be transferred to an intensive care unit, including nine in the prolonged-hospitalization group. In total, 81 patients (87.1%) were discharged in a better condition, 7 patients (7.5%) died in hospital, 3 patients (3.2%) were transferred to another health care setting and 2 patients (2.2%) refused to stay in hospital. The prolonged-stay group had a higher in-hospital mortality than the shorter-stay group, but this was not statistically significant.

Table 4. Treatment-related factors, complications and hospital outcomes

The variables with a p value < 0.20 in the univariate analysis were included in the multivariable logistic regression model. The number of comorbidities, the presence of sepsis on admission, development of an in-hospital complication and hypoalbuminaemia showed a strong association with prolonged hospitalization, which remained significant after adjustments were made to the potential confounding factors. For every patient with 2 or more comorbidities, the odds of extended hospitalization were roughly 3 times greater than if a patient had fewer comorbidities. Sepsis at admission was associated with more than four times the odds of extended hospital stay as was acute kidney injury (nearly four times). A serum albumin level <3.5 g/dL and the presence of an in-hospital complication were also good independent predictors of hospital duration of >7 days.

Table 5. Multivariable logistic regression analysis of predictors of prolonged hospitalization

In summary, prolonged hospitalization occurred in slightly more than one-third of patients admitted to the internal medicine units. Multimorbidity, sepsis, acute kidney injury, hypoalbuminaemia and in-hospital complications were identified as the principal independent predictors of prolonged hospital stay.

Figure 1. Distribution of patients according to duration of hospitalization. Among 93 patients, 59 (63.4%) had a hospital stay of 7 days or less, whereas 34 (36.6%) experienced prolonged hospitalization of more than 7 days.

The present prospective cohort study found that 34 of 93 patients (36.6%) admitted to the internal medicine units experienced prolonged hospitalization, defined as a hospital stay exceeding 7 days. Patients with prolonged hospitalization were generally older and had a greater burden of comorbid diseases, more severe clinical abnormalities at admission and a higher frequency of complications during their hospital course. After adjustment for possible confounding variables, multimorbidity, sepsis at admission, acute kidney injury, hypoalbuminaemia and the development of an in-hospital complication remained independent predictors of prolonged hospitalization. These findings support the understanding that length of stay among medical inpatients is influenced by a combination of baseline vulnerability, acute disease severity, physiological reserve and events occurring during hospitalization. Previous studies involving general medical populations have similarly shown that a relatively small group of patients with prolonged stays may consume a disproportionately large number of inpatient bed-days and hospital resources [8, 9]. A systematic review published in 2022 also emphasized that reliable prediction of hospital length of stay requires the integration of demographic, clinical, laboratory and hospital-process variables rather than dependence on a single factor [6]. In the current study, one of the important independent factors associated with long hospital stays was multimorbidity. The odds of being hospitalized for over 7 days were around three times higher for patients with two or more comorbid conditions. The presence of multiple chronic illnesses can make diagnostic evaluation more complicated, result in conflicting treatment priorities and make it more difficult to recover from an acute illness. These patients often need further investigations, changes in medication, referral to specialists and multidisciplinary care and discharge planning. The strong link between resource use and cumulative disease burden is further supported by the significant difference in Charlson Comorbidity Index between the patient groups with prolonged hospitalization and those without. In a large study of medical inpatients in Switzerland it was found that a multimorbidity was correlated with longer hospital stays, higher intensive care unit utilization and poorer hospital outcomes [10]. In another multi-national study, including over 147,000 medical inpatients, multimorbidity and the burden of comorbidities were found to be good predictors of length of stay and early readmission [11]. A 2022 systematic review and meta-analysis also found a connection between multimorbidity in older adults and higher hospitalizations and healthcare utilization [12]. Other independent factors of prolonged hospitalisation were sepsis and acute kidney injury. The adjusted OR for prolonged hospitalization was over 4 times greater for patients with sepsis, and nearly 4 times greater for patients with acute kidney injury. Typical complications of sepsis include systemic inflammation, haemodynamic instability, tissue hypoperfusion and dysfunction of one or more organs. Long-term intravenous antimicrobial therapy, fluid resuscitation, frequent laboratory monitoring, oxygen support, and in the worst cases, the use of vasopressors and intensive care may be required in these patients. Fluid and electrolyte disturbances, the need for changes in medication doses and renal replacement therapy (RRT) due to uraemic complications can further delay recovery from acute kidney injury. Previous larger studies found that the incidence of AKI in hospitalized patients was correlated with a higher death rate, longer hospital stay and significant higher treatment cost [13, 14]. Within this current cohort, other parameters associated with prolonged stays were also more common in patients who had elevated creatinine, hyponatraemia, increased C-reactive protein and required oxygen, suggesting a role of organ dysfunction and severity of systemic illness in a longer duration of stay. Another independent predictor of long hospitalization was hypoalbuminaemia. Adjusted odds of having a duration of hospital stay greater than 7 days were significantly higher in patients with a serum albumin below 3.5 g/dL (about 4 times higher). Nutritional status, liver production, inflammation, renal or gastrointestinal losses, and capillary leakage in acute illness all affect the serum albumin. Low albumin levels can be interpreted as a sign of chronic frailty and also of the magnitude of an acute inflammatory response. There was a meta-analysis that included 90 cohort studies, which showed the association with declining serum albumin and increased morbidity, mortality, resource utilisation and hospitalisation time [15]. Likewise, hypoalbuminaemia was found to be associated with higher comorbidity, renal dysfunction, anaemia, infection and worse outcomes among patients with internal medicine [16, 17]. A study of hospitalized medical patients in 2022 revealed that medical patients with hypoalbuminaemia had a higher Charlson Comorbidity Index score, a longer median hospital stay and increased hospital readmissions compared to those with normal albumin levels [18]. These observations are similar to the current observations and indicate that the serum albumin, which is inexpensive and routinely available, may be useful for recognizing those patients who need early nutritional evaluation, increased monitoring and coordinated discharge planning. Development of an in-hospital complication was the best predictor in the final regression model, which more than doubled the adjusted odds of length of hospital stay. Patients staying in hospital longer were more likely to have a hospital acquired infection. This dynamic can be two-way: the length of the hospital stay can lead to exposure to invasive devices, resistant microorganisms, and decreased mobility; conversely, a hospital-acquired complication can result in needing additional treatment and further delay of hospital discharge. Large hospital-based studies have demonstrated that complications that occur during the stay are correlated with significant extension of hospital stay and cost of treatment [19, 20]. The present findings underscore the importance of early identification of high-risk patients, infection control measures, and catheter care, pressure-injury prevention, early mobilisation, medication checks and regular evaluation of patients to ensure the absence of patients-specific discharge barriers. The results of this study, however, must be interpreted in the light of the small number of patients, single centre study, and short study period. The use of a single threshold of > 7 days may also reduce the ability to compare with studies that have adopted either the upper quartile (14 days or 21 days) as the threshold for prolonged hospitalization. These predictors need to be confirmed in larger multicentre studies and a simple measure based on admission that can be used in an internal medicine unit should be created.

Prolonged hospitalization occurred in more than one-third of patients admitted to the internal medicine units. The principal independent predictors were the presence of two or more comorbidities, sepsis at admission, acute kidney injury, hypoalbuminaemia and the development of an in-hospital complication. These findings indicate that prolonged hospitalization is associated with both the patient’s underlying health status and potentially modifiable clinical events occurring during admission. Early risk assessment using comorbidity burden, renal function, serum albumin and indicators of acute illness severity may allow healthcare teams to identify patients likely to require extended care. Timely treatment of sepsis and acute kidney injury, nutritional assessment, prevention of hospital-acquired complications and early multidisciplinary discharge planning may help reduce avoidable hospital days and improve the use of limited inpatient resources.