The length of hospital stay after acetaminophen overdose

BACKGROUND: Acetaminophen is one of the most commonly encountered medications in self-poisoning, with a high rate of morbidity. The prevalence and characteristics of acetaminophen intoxication associated with long hospital stay in patients are not well defined.

OBJECTIVE: This study aims to identify the clinical and demographic factors associated with the length of in-hospital stay (LOS), and to evaluate the effectiveness of early treatment of acetaminophen overdose patients (≤ 8 hours) by intravenous N-acetylcysteine (IV-NAC) on hospital stay.

METHODS: This is a retrospective cohort study of hospital admissions for acetaminophen overdose conducted over a period of five years from 1 January 2004 to 31 December 2008. Patients were divided into two groups: LS group patients had a long hospital stay (> median hours stay in hospital) and SS group patients had a short hospital stay (≤ median hours stay in hospital). SPSS 15 was used for data analysis.

Get Help With Your Nursing Essay

If you need assistance with writing your nursing essay, our professional nursing essay writing service is here to help!

Nursing Essay Writing Service

RESULTS: Of the 305 patients, eleven significant factors were identified in the univariate analysis as significantly associated with LS. Three independent factors were found to be significant predictors of LS in the multivariate analysis. The significant factors associated with LS were seen among patients with a history of abdominal pain after ingestion of acetaminophen (P = 0.04), who were on IV-NAC administration (P < 0.001), and had an acutely depressed mood (P = 0.003). Significantly late time to NAC infusion of more than 8 hours was associated with LS rather than SS (96 patients (57%) and 6 (24%), respectively; P = 0.003).

CONCLUSION: Patients with long hospital stay have different clinical characteristics compared to patients with short hospital stay. We identified IV-NAC administration as a potentially modifiable factor that may lead to prolonged hospital stay. When risk assessment indicates that NAC is required, it is highly recommended that NAC be started in the first hours of admission to reduce the length of hospital stay.

Key Words: Acetaminophen; Overdose; Hospital stay; N-acetylcysteine.

Introduction

Acetaminophen (paracetamol) is a commonly used analgesic and antipyretic drug [1]. Acetaminophen overdose has been extensively reported in the USA and UK [2, 3]. It remains the most common means of pharmaceutical poisoning in the eastern nations, including Malaysia [4, 5].

Hepatotoxicity and nephrotoxicity are the main characteristic features of acetaminophen poisoning [6, 7]. The risk of toxicity is initially determined from the extent of acetaminophen exposure after considering the stated amount ingested and comparison of serum acetaminophen concentrations to the Rumack-Matthew nomogram [8, 9]. The extent of hepatic and renal injury is later determined by measurement of serum liver enzyme activity, prothrombin time, and creatinine concentrations [10].

For adults, ingestion of 150 mg/kg or more [11.5 g for a 70 kg adult (23 tablets of 500 mg)] should be considered potentially hepatotoxic [11]. Management of poisoned patients includes activated charcoal within the first hour after ingestion and administration of the antidote N-acetylcysteine (NAC), which reduces complications and provides almost complete protection against liver necrosis if given within 8-10 hours of acetaminophen ingestion [12, 13].

Factors influencing the length of in-hospital stay (LOS) in poisoned patients have received less attention in the literature, and there are no previously published studies regarding the LOS after acute acetaminophen overdose. Determining the factors influencing the hospitalization period may help decrease the LOS, reduce costs, and improve the efficiency of management.

To improve our knowledge about hospital stay after acute acetaminophen overdose, we carried out this three-year hospital-based study with the following objectives: 1) to investigate the prevalence of prolonged hospital stay in patients admitted to hospital after acute acetaminophen overdose, 2) to identify the clinical and demographic factors associated with prolonged hospital stays, and 3) to evaluate the effectiveness of early treatment of acute acetaminophen overdose patients (≤ 8 hours) by intravenous N-acetylcysteine (IV-NAC) on prolonged hospital stays. The knowledge of prevalence, clinical characteristics, and predictors of prolonged hospital stay in patients after acetaminophen overdose might contribute to reducing complication rates by enhancing the application of specific therapeutic and management strategies to patients at high risk of hepatotoxicity. Furthermore, knowledge of predictors in the first hours of hospital admission still deserves special attention because even a quantitative decrease in hospital stay may be translated into substantial savings in health care costs and resources.

Methods

Settings and Study Design

This is an observational retrospective case review of all patients with acute acetaminophen overdose admitted to a 1200-bed hospital located in the northern region of Malaysia. The hospital provides health care and emergency treatment for all illnesses and accidents. All aspects of the study protocol, including access to and use of the patient clinical information, were authorized by the local health authorities before initiation of this study.

Participants and Data Collection

Data were collected during the period of study from 1 January 2004 to 31 December 2008. A computer generated list was obtained from the Hospital Record Office. We identified our cases according to the T-codes of the International Classification of Diseases tenth revision (ICD-10). All patients with diagnostic code T 39.1 (acetaminophen overdose) were included in the study. Patients’ medical records were reviewed systematically. Specially designed data-collection forms were used to collect data. They were divided into two groups: an LS group with long hospital stays and an SS group with short hospital stays. Long hospital stay was defined as a length of hospital stay greater than the median. Length of stay was calculated as the hour of discharge minus the hour of admission.

Variables were documented from medical records for comparison between the two groups. A total of 20 variables were identified for comparison. These included nine patient variables (age, gender, ethnicity, cause of overdose including intentional or unintentional, history of psychiatric illness, history of chronic illness, history of alcohol intake, history of suicide attempt, and acute depressed mood) and eleven acetaminophen variables (number of ingested agents, post-ingestion nausea, post-ingestion vomiting, post-ingestion diarrhoea, post-ingestion abdominal pain, reported dose ingested, the latency time (the time of ingestion to the time the patient was presented at the hospital), stomach wash done, activated charcoal intake, NAC administration, and estimated acetaminophen level according to whether it was above or below the ‘high-risk’ treatment line as stated in the Rumack-Matthew nomogram (≥ 150 mg/L or < 150 mg/L [8])). Acute depressed mood was defined as the presence of causes of deliberate self-harm such as depression, anxiety, and adjustment disorders; these causes were noted by the hospital psychiatric specialist report.

We therefore studied all patients who were treated with intravenous N-acetylcysteine to determine whether a long hospital stay was more likely to occur in patients with late time to NAC administration (> 8 hours). Two categories were used for time from acetaminophen ingestion to IV-NAC administration: ≤ 8 and > 8 hours.

Statistical Analysis

Data were entered and analysed using the Statistical Package for Social Sciences program version 15 (SPSS). Data were expressed as means ± SD for continuous variables and as frequencies for categorical variables. Variables that were not normally distributed were expressed as medians (lower-upper quartiles). Variables were tested for normality using the Kolmogorov-Smirnov test. Either the Chi square or the Fisher exact test, as appropriate, was used to test significance between categorical variables. The student’s t test was used to compare the means of continuous variables. If assumptions of equality of variance and normality (assumed for the t test) were not met, the Mann-Whitney U test (a nonparametric equivalent of the t test) was performed as appropriate. Multiple logistic regression analysis was used to identify factors associated with a length of stay greater than the median. Variables included in the regression were those with significant P values (< 0.05) in the univariate analysis. The proportion of patients with long hospital stays in the study group was expressed as a prevalence rate with a 95% confidence interval (95% CI). The association between LS group and SS group and the variables of interest was evaluated by calculating the odds ratio (OR) with the corresponding 95% CI.

Results

Three hundred and five cases of acetaminophen overdose were identified. Table 1 shows the demographics and clinical characteristics of acetaminophen poisoning cases. Two hundred and fifty-six (83.9%) of the cases were females and forty-nine (16.1%) were males, giving a male: female ratio of 0.19: 1. The average age of the cases was 23.07 ± 7.34. Initial management included a stomach wash, which was performed in 180 cases (59%). Activated charcoal was given while patients were in the Accident and Emergency department. It was given as single or multiple doses in 190 cases (62.3%). Intravenous NAC was given in 146 cases (47.9%) after estimation of acetaminophen levels. Overall, two patients were admitted to the intensive care unit but no patient died or needed a liver transplant as a result of acetaminophen overdose. Also, only two patients with acetaminophen overdose were presented to the hospital with impairment in level of consciousness upon admission.

The distributions of LOS in the study population are shown in Figure 1. Across the entire study period, the median (interquartile range) LOS was 36 hours (20-61 hours). Thus, an increased length of stay was defined as longer than 36 hours. Of the 305 acute acetaminophen overdose cases, 156 patients (51.1%) had long hospital stays. The medians (interquartile ranges) were 20 hours (16-26 hours) and 62 hours (46-80.75 hours) in the SS and LS groups, respectively.

Find Out How NursingAnswers.net Can Help You!

Our academic experts are ready and waiting to assist with any writing project you may have. From simple essay plans, through to full dissertations, you can guarantee we have a service perfectly matched to your needs.

View our academic writing services

The results of univariate analysis in 305 patients according to demographic and clinical characteristics status on admission are shown in Table 1. Eleven significant factors were identified by the comparison of 20 variables between the SS and LS groups in univariate analysis. Long hospital stay was significantly associated with intentionally ingested drug (OR = 1.71; 95% CI = 1.1-2.65; P = 0.005), patients with acute depressed mood (OR = 2.32; CI = 1.73-3.11; P < 0.001), reported acetaminophen dose ingested ≥ 10 grams (OR = 3.03; CI = 2.16-4.23; P < 0.001), latency time of more than 8 hours (OR = 1.57; CI = 1.23-1.95; P < 0.001), patients without stomach wash (OR = 1.38; 95% CI = 1.1-1.73; P = 0.005), patients without activated charcoal intake (OR = 1.28; 95% CI = 1.02-1.6; P = 0.037), NAC administration (OR = 4.71; 95% CI = 3.34-6.64; P < 0.001), patients with abdominal pain at presentation (OR = 1.64; 95% CI = 1.32-2.03; P < 0.001), patients with vomiting at presentation (OR = 1.49; 95% CI = 1.12-1.95; P = 0.003), patients with diarrhoea at presentation (OR = 1.79; 95% CI = 1.29-2.47; P = 0.038), and patients with acetaminophen level above the ‘high-risk’ treatment line (OR = 3.58; 95% CI = 2.69-4.75; P < 0.001).

Table 2 shows the multivariate logistic regression analysis of factors related to long hospital stay. Three significant factors were identified in the comparison of 11 variables between the SS and LS groups in multivariate logistic regression analysis. Multiple logistic regression showed that significant factors for long hospital stay were associated with patients who had abdominal pain at presentation (P = 0.04), patients who had acute depressed mood (P = 0.003), and NAC administration (P < 0.001). The model was significant with a Chi-square of 173.5, DF = 11; P < 0.001.

We further analysed and compared the time to NAC infusion associated with length of in-hospital stay. As shown in Figure 2, the median (interquartile range) times to NAC infusion in the LS and SS groups were 9 hours (6-12.5 hours) and 6 hours (5.25-9 hours), respectively. A significantly late time to NAC infusion of more than 8 hours was associated with long hospital stay rather than short hospital stay (96 patients (57%) and 6 (24%), respectively; P = 0.003)

Discussion

The present study was conducted to investigate the prevalence of prolonged hospital stay in patients presenting to hospital, to identify the clinical and demographic factors associated with prolonged hospital stay, and to evaluate the effectiveness of early treatment of acute acetaminophen overdose patients (≤ 8 hours) by IV-NAC on prolonged hospital stays in 305 consecutive patients with acute acetaminophen overdose collected retrospectively from the records registry.

In this study, the hospital duration ranged from 4 hours to 12.5 days (mean: 46 hours). A previous study found that hospital duration ranged from 20 minutes to 111 days (mean: 26 hours) [14]. In a previous retrospective study with 93 patients presenting to the emergency department after acetaminophen overdose, the mean hospital stay was 4.25 days. The hospital stay was higher in the accidental overdose group (mean 6.4 ± 6.1 days) in comparison to the suicidal overdose group (3.9 ± 2.7 days) [15].

In our study, patients with poisoning due to acetaminophen ingestion had different hospital stays in comparison to previous studies. These observed differences in hospital stay may be due in part to the quality of medical care and patient factors. These findings still deserve a special attention because a quantitative decrease in hospital stay may be translated into substantial savings in health care costs and resources [16].

In the current study, having a history of abdominal pain after ingestion of acetaminophen, IV-NAC administration, and an acutely depressed mood were identified as predictors of long hospital stay.

Our study has found that a history of abdominal pain after ingestion of acetaminophen was an independent predictor for long hospital stay. To our knowledge, abdominal pain is considered one of clinical signs suggestive of acetaminophen toxicity [12]. A previous study has been conducted to identify risk markers for hepatic failure in acetaminophen overdose [17]. In that brief report, Knell listed abdominal pain as associated with hepatic failure [17]. This suggests that early evaluation of serum acetaminophen and alanine aminotransferase (ALT) and early NAC administration among these patients with abdominal pain have the potential to improve the outcome after acetaminophen overdose [12].

Our data showed that LOS was significantly more common in patients with NAC administration. The high prevalence of long hospital stay after acetaminophen overdose might be due to delay in NAC administration after estimation of acetaminophen levels among patients with high reported ingested doses. The current study observed that a long hospital stay was associated with a late time to NAC infusion of more than 8 hours. The delay from acetaminophen ingestion to antidote (NAC) treatment deserves close attention. N-acetylcysteine administration was the second most important independent risk factor in the model. A Danish study showed that all patients developing hepatic failure had a delay of at least 18 hours before receiving antidote treatment [18]. These findings contrast with a British report on 560 patients with acetaminophen intoxication, in which 20% of those treated within 12 hours after overdose died [19]. A point of concern when using the antidote delay as a causal factor for LOS is that physicians may sometimes have difficulties obtaining an accurate and reliable history with regard to the exact time of ingestion or amount of acetaminophen ingested. This could be due to the emotional state, forgetfulness, or direct misleading by the patient. Despite these possible biases the amount of acetaminophen ingested and latency time proved valuable in the univariate model, and this information should be used in a ‘positive’ manner: a delayed time to antidote should increase the attention to the clinical status of the patient.

The mean time from exposure to treatment has been reported to significantly affect the prognosis and final outcome of acetaminophen toxicity. A panel of Australian and New Zealand clinical toxicologists stated that for patients who present 8 or more hours after ingestion, NAC should be commenced immediately if the reported dose exceeds the threshold for possible toxicity or the patient shows clinical signs suggestive of acetaminophen toxicity (nausea, vomiting, right upper quadrant pain or tenderness). Evaluation of serum acetaminophen and ALT levels should then be performed as soon as possible. If the serum acetaminophen level is subsequently found to be below the nomogram line, NAC may be ceased; if above the line, it should be continued. The baseline serum ALT level assists risk assessment and provides useful baseline data if NAC is indicated. Similarly, if NAC is commenced, the baseline international normalized ratio and platelet count provide additional data to inform later risk assessments (e.g., for risk of death from hepatic failure) [12].

For patients who present within 8 hours with a known time of ingestion, risk assessment is based on the serum acetaminophen level plotted on the nomogram. Supplementary investigations such as liver function tests or a coagulation profile do not refine the risk assessment, and do not provide useful baseline data or change management in this group of patients. These tests are therefore not indicated unless risk assessment for another agent requires them. Follow-up tests are not required at the conclusion of the 20-hour NAC infusion, before discharge, or at subsequent follow-up [12, 20].

In the present study, acute depressed mood was an independent predictor of LOS. Several studies conducted in general hospitals have tried to determine the impact of psychiatric illness on LOS with conflicting results [20-26]. The main difficulty is evaluating whether having a psychiatric illness is an independent risk factor for increased LOS, and if so, which specific mental disorders can influence LOS. A critical review of 26 outcome studies that examined the association of psychiatric illness and LOS in the general hospital showed that impaired cognition associated with delirium and dementia, depressed mood, and other personality variables contribute to prolonged hospital stays [27]. In addition, it must be considered that LOS may be influenced in addition by the health care system itself. To our knowledge, comparative epidemiological studies investigating the influence of psychiatric morbidity in patients with acetaminophen overdose on LOS in different health care systems and use of a prospective design and comparable instruments have not been done until now. For this reason, we do not know whether our results are true for countries with health care systems different from that in Malaysia. It is recommended that patients suffering a serious psychiatric disorder and/or at high risk of suicide are likely to require in-patient treatment of the underlying psychiatric disorder [28]. In a previous study, the authors advised that all patients presenting to hospital with an episode of self-harm should have a psychosocial assessment [29]. This is because of the significant risk of completed suicide following deliberate self-poisoning and the fact that approximately half of all suicides in the United Kingdom have a history of deliberate self-harm [30].

Conclusion and Recommendations

Patients with long hospital stays have different clinical characteristics compared to patients with short hospital stays. Our data suggest that abdominal pain at presentation, presence of psychiatric illness, and delay in NAC administration were associated with prolonged hospital stay. We identified NAC administration as a potentially modifiable factor that may lead to prolonged hospital stays. When risk assessment indicates that NAC is required primarily from the stated quantity of ingested acetaminophen, it is highly recommended that NAC should be started during the first hours of admission. This potentially modifiable factor would be an ideal target for studies aiming to reduce the burden of illness and healthcare costs of acute acetaminophen overdose, because LOS is a major cost component of hospital budgets. This highlights the need for increased education for physicians who work in the emergency room on the proper care of patients who present with acute acetaminophen [20]. In many hospitals, the emergency room, especially at night, is staffed with residents or young physicians with little formal training in toxicology [31]. In addition, all accident and emergency department staff who manage DSH patients need to know how to conduct a brief psychosocial assessment, especially with regard to identifying risk factors for repetition of suicide attempts and for suicide [29].

Acknowledgments: The authors would like to thank Universiti Sains Malaysia (USM) for the financial support provided for their research. The assistance of the medical and record office staff is gratefully acknowledged.

Conflict of interests: We would like to declare that there is no conflict of interests in conducting this research.