Risk factors and outcomes of peptic ulcer bleed in a Pakistani population: A single-center observational study

Abstract

BACKGROUND

Peptic ulcer disease (PUD) remains a significant healthcare burden, contributing to morbidity and mortality worldwide. Despite advancements in therapies, its prevalence persists, particularly in regions with widespread nonsteroidal anti-inflammatory drugs (NSAIDs) use and Helicobacter pylori infection.

AIM

To comprehensively analyse the risk factors and outcomes of PUD-related upper gastrointestinal (GI) bleeding in Pakistani population.

METHODS

This retrospective cohort study included 142 patients with peptic ulcer bleeding who underwent upper GI endoscopy from January to December 2022. Data on demographics, symptoms, length of stay, mortality, re-bleed, and Forrest classification was collected.

RESULTS

The mean age of patients was 53 years, and the majority was men (68.3%). Hematemesis (82.4%) and epigastric pain (75.4%) were the most common presenting symptoms. Most patients (73.2%) were discharged within five days. The mortality rates at one week and one month were 10.6% and 14.8%, respectively. Re-bleed within 24 h and seven days occurred in 14.1% and 18.3% of patients, respectively. Most ulcers were Forrest class (FC) III (72.5%). Antiplatelet use was associated with higher mortality at 7 and 30 d, while alternative medications were linked to higher 24-hour re-bleed rates. NSAID use was associated with more FC III ulcers. Re-bleed at 24 h and 7 d was strongly associated with one-week or one-month mortality.

CONCLUSION

Antiplatelet use and rebleeding increase the risk of early mortality in PUD-related upper GI bleeding, while alternative medicines are associated with early rebleeding.

Key Words: Non variceal bleed; Mortality; Re-bleed; Forrest classification; Antiplatelets; Alternative medicines

Core Tip: This Pakistani study flags antiplatelets and alternative meds as risk factors for peptic ulcer bleeding mortality and re-bleed. Considering patient comorbidities and tailoring treatment based on these factors, like avoiding antiplatelets if possible, could improve outcomes in this population. However, larger studies are needed to solidify these findings and personalize treatment further.

INTRODUCTION

Non-variceal upper gastrointestinal (UGI) bleed is an acute medical emergency and a common cause of hospital admission[1]. Peptic ulcer disease (PUD) related bleeding is one of the major causes of non-variceal UGI bleeding and is a major player in morbidity and mortality[2]. Despite all the advances in Proton pump inhibitor (PPI) therapy, increasing population age along with widespread use of nonsteroidal anti-inflammatory drugs (NSAIDs) and Helicobacter pylori (H. pylori) in our part of the world, PUD has a major share in gastrointestinal (GI) related morbidity and mortality. Its estimated prevalence is around 5%-10%[3]. Infection with H. pylori has long been associated with peptic ulcer formation[4]. The increasing population age with ever-increasing use of steroids, NSAIDs, Aspirin, and smoking are other risk factors[5-7]. As expected, the combined use of Corticosteroids along with antiplatelets or anticoagulants increases the risk of bleed from Peptic ulcers even further[8]. Although the Forrest classification is a widely used tool for predicting re-bleed risk in peptic ulcer bleeding (PUB), evidence suggests that comorbid conditions significantly impact outcomes[9,10]. Re-bleed, length of stay (LOS) and mortality are important outcome parameters to evaluate in cases of PUB. The burden of the disease, along with these outcome parameters, is supposed to have improved with the improvement in endoscopy and overall healthcare facilities. However, despite advances in treatment, PUB remains a significant health burden, and the mortality has remained at 5%-10% over the past three decades[11-14], particularly in regions like Pakistan with high NSAID use and H. pylori prevalence[15].

Existing research on PUB outcomes primarily comes from Western populations, potentially overlooking regional factors and specificities. This study aims to fill knowledge gaps by comprehensively analyzing risk factors and outcomes of PUB in a Pakistani population, providing insights relevant to local healthcare practices. This study further delves deeper by investigating the role of comorbidities in re-bleed, mortality, and LOS, as well as the association between re-bleed and mortality.

MATERIALS AND METHODS

This retrospective cohort study was conducted at the Gastroenterology Department of Jinnah Post Graduate Medical Centre, Karachi, from January 2022 to December 2022. A total of 1098 patients were presented to the unit with suspected upper GI bleed (hematemesis or melena). Among them, 142 patients had confirmed peptic ulcer bleed, which were included in the study. Other causes of GI bleed included variceal bleed (702), esophageal cancer (11), gastric malignancies (53), lower GI bleed (155), Mallory Weiss tear (7) and Gastric Antral Vascular Ectasia (28). Patients with other causes of GI bleed and incomplete data were excluded from the study. The detailed demographics were recorded on pre-designed proformas. Any specific medical condition/comorbidity or drugs such as antiplatelets, NSAIDs, Corticosteroids or alternative (herbal/homoeopathic/Hakimi) medications were also extracted from medical records.

The vitals of the patients, along with the presence or absence of shock or tachycardia, were meticulously recorded. The bleeding control pathway was used as a guide for therapy during the hospital stay. Patients were started on intravenous PPIs and were admitted to an appropriate care level after proper triage. Each patient was assessed individually according to the American Society of Anesthesiology for anaesthesia risk in relation to comorbidity. Endoscopy findings were recorded and documented according to Forrest classification.

All patients who were included in the study underwent endoscopy during the initial 24 h as dictated by the international guidelines[1]. An urgent endoscopy was considered among those patients whose bleed was severe enough to exclude medical management pre-procedure. A peptic ulcer was defined as a breach in the mucosal integrity on endoscopy. Further details of the ulcer, such as its location (gastric or duodenal) and its severity according to the Forrest classification, were recorded through the endoscopy report. The Forrest classification refers to the endoscopic appearance of a peptic ulcer as active spurting (Forrest IA), active oozing (Forrest IB), nonbleeding visible vessel (Forrest IIA), ulcers with adherent clots (Forrest IIB), ulcers with red spots (Forrest IIC) or a clean base (Forrest III). Patients with high-risk ulcers (Forrest 1 and 2) were treated endoscopically with at least dual endoscopic interventional modalities (Sclerotherapy along with hemo-clips).

All patients who received dual endoscopic modality to achieve hemostasis were admitted to the High Dependency Unit post- procedure and were closely monitored for re-bleed. Their vitals and laboratory parameters were recorded according to the bleeding pathway protocol. Target Hemoglobin (Hb) was maintained around 8-9 gm/dL, depending upon their cardiac and volume status.

Any rebleed manifested in the form of Malena, drop in Hb, and/or endoscopic evidence of upper gastrointestinal bleed (UGIB) within 24 h or one week was also recorded. The primary outcomes were LOS, rebleeding in 24 h and one week, and mortality in one week and one month. This study was conducted after approval from the Institutional Review Board of the Jinnah Postgraduate Medical Centre, Karachi.

Statistical analysis

Descriptive statistics were reported as frequency (percentage) for categorical variables and mean ± SD for continuous variables. The association between outcome variables and risk factors was assessed using the chi-square test/Fisher's exact, where appropriate. A P value < 0.05 was considered significant. SPSS 27 was used for statistical analyses.

RESULTS

A total of 142 patients with UGIB secondary to PUD were included in the study. The mean age of the patients was 53 ± 18.5 years, including 97 (68.3%) males and 45 (31.7%) females. The most common presenting symptom was hematemesis and epigastric pain among 117 (82.4%) and 107 (75.4%) patients, respectively. Common comorbid conditions included hypertension (37.3%), diabetes (25.4%) and ischemic heart disease (26.8%; Table 1).

Table 1 Demographic and clinical characteristics of patients admitted to hospital with upper gastrointestinal bleed (n = 142).

Characteristics	n	%
Age in years, mean ± SD (min-max)	53.0 ± 18.5	18-90
Gender
Male	97	68.3
Female	45	31.7
Comorbidity
Diabetes mellitus	36	25.4
Hypertension	53	37.3
Ischemic heart disease	38	26.8
Asthma	7	4.9
Cerebrovascular accident	14	9.9
Presenting symptoms
Epigastric pain	107	75.4
Bloating	30	21.1
Hematemesis	117	82.4
Melena	88	62.0
Abdominal pain	19	13.4
Weight loss	12	8.5
Location of ulcer
Stomach	54	38.0
Duodenum	63	44.4
Stomach and duodenum	25	17.6

The hospital stay in most patients was ≤ 5 d (n = 104, 73.2%). Fifteen (10.6%) patients died in the first week, while 21 died in one month (14.8%). Re-bleed within 24 h occurred among 20 (14.1%) patients, while 26 (18.3%) patients presented with recurrent bleeding in 7 d.

Those patients who were on antiplatelet drugs, the proportion of 7 d mortality (21.3% vs 5.3%, P = 0.007) and 30 d mortality (23.4% vs 10.5%, P = 0.042) was high as compared to those who were not on antiplatelet drugs. For those who were on alternative medications, re-bleed in 24 h was high (21.4% vs 9.3%, P = 0.042). Smoking, steroids and H. pylori did not appear to impact LOS, rebleed, and mortality (Table 2).

Table 2 Association of risk factors and forest classification with outcome variables.

Variables	Total, n (%)	Hospital stay in days		Mortality in 7 d		Mortality in 30 d		Re-bleed in 24 h		Re-current bleed in 7 d
		≤ 5 (n = 104)	> 5 (n = 38)	No (n = 127)	Yes (n = 15)	No (n = 121)	Yes (n = 21)	No (n = 122)	Yes (n = 20)	No (n = 116)	Yes (n = 26)
Smoking
Yes	31 (21.8)	24 (77.4)	7 (22.6)	29 (93.5)	2 (6.5)	29 (93.5)	2 (6.5)	27 (87.1)	4 (12.9)	26 (83.9)	5 (16.1)
No	111 (78.2)	80 (72.1)	31 (27.9)	98 (88.3)	13 (11.7)	92 (82.9)	19 (17.1)	95 (85.6)	16 (14.4)	90 (81.1)	21 (18.9)
Steroids
Yes	8 (5.6)	7 (87.5)	1 (12.5)	7 (87.5)	1 (12.5)	7 (87.5)	1 (12.5)	7 (87.5)	1 (12.5)	7 (87.5)	1 (12.5)
No	134 (94.4)	97 (72.4)	37 (27.6)	120 (89.6)	14 (10.4)	114 (85.1)	20 (14.9)	115 (85.8)	19 (14.2)	109 (81.3)	25 (18.7)
H. pylori
Yes	26 (18.3)	20 (76.9)	6 (23.1)	24 (92.3)	2 (7.7)	23 (88.5)	3 (11.5)	22 (84.6)	4 (15.4)	20 (76.9)	6 (23.1)
No	116 (81.7)	84 (72.4)	32 (27.6)	103 (88.8)	13 (11.2)	98 (84.5)	18 (15.5)	100 (86.2)	16 (13.8)	96 (82.8)	20 (17.2)
Antiplatele
Yes	47 (33.1)	34 (72.3)	13 (27.7)	37 (78.7)	10 (21.3)b	36 (76.6)	11 (23.4)a	37 (78.7)	10 (21.3)	35 (74.5)	12 (25.5)
No	95 (66.9)	70 (73.7)	25 (26.3)	90 (94.7)	5 (5.3)	85 (89.5)	10 (10.5)	85 (89.5)	10 (10.5)	81 (85.3)	14 (14.7)
Alternative medications
Yes	56 (39.4)	39 (69.6)	17 (30.4)	47 (83.9)	9 (16.1)	44 (78.6)	12 (21.4)	44 (78.6)	12 (21.4)a	43 (76.8)	13 (23.2)
No	86 (60.6)	65 (75.6)	21 (24.4)	80 (93.0)	6 (7.0)	77 (89.5)	9 (10.5)	78 (90.7)	8 (9.3)	73 (84.9)	13 (15.1)
Forrest classification
I	18 (12.7)	10 (55.6)	8 (44.4)	18 (100.0)	0	17 (94.4)	1 (5.6)	13 (72.2)	5 (27.8)	11 (61.1)	7 (38.9)a
II	21 (14.8)	14 (66.7)	7 (33.3)	18 (85.7)	3 (14.3)	17 (81.0)	4 (19.0)	17 (81.0)	4 (19.0)	16 (76.2)	5 (23.8)
III	103 (72.5)	80 (77.7)	23 (22.3)	91 (88.3)	12 (11.7)	87 (84.5)	16 (15.5)	92 (89.3)	11 (10.7)	89 (86.4)	14 (13.6)

^aP < 0.05.

^bP < 0.01.

n (%) are reported. P value was calculated by Chi-square/Fisher Exact test. H. pylori: Helicobacter pylori.

On endoscopy, most of the ulcers belonged to Forrest class (FC) III (n = 103, 72.5%), followed by FC II (n = 21, 14.8%) and FC I (n = 18, 12.7%). Patients with FC I showed a higher proportion of re-current bleeding in 7 d (38.9% vs 23.8% vs 13.6%, P value = 0.034) than those with FC II and FC III (Table 2). However, FC was not found to be significantly associated with any of the risk factors (Table 3).

Table 3 Association of risk factors with Forest classification.

Risk factors	Total	Forrest classification
		I (n = 18)	II (n = 21)	III (n = 103)
Smoking
Yes	31 (21.8)	2 (11.1)	6 (28.6)	23 (22.3)
No	111 (78.2)	16 (88.9)	15 (71.4)	80 (77.7)
Steroids
Yes	8 (5.6)	0	1 (4.8)	7 (6.8)
No	134 (94.4)	18 (100.0)	20 (95.2)	96 (93.2)
H. pylori
Yes	26 (18.3)	5 (27.8)	5 (23.8)	16 (15.5)
No	116 (81.7)	13 (72.2)	16 (76.2)	87 (84.5)
Antiplatelet
Yes	47 (33.1)	5 (27.8)	7 (33.3)	35 (34.0)
No	95 (66.9)	13 (72.2)	14 (66.7)	68 (66.0)
Hakeemi medications
Yes	56 (39.4)	5 (27.8)	7 (33.3)	44 (42.7)
No	86 (60.6)	13 (72.2)	14 (66.7)	59 (57.3)

n (%) are reported. P value was calculated by Chi-square/Fisher Exact test. H. pylori: Helicobacter pylori.

It was noted that the patients who used NSAIDs had more with FC III (76.1% vs 58.6%, P value = 0.037) as compared to those who did not report the use of NSAIDs (Table 4). Re-bleed in 24 h and seven days was found to be statistically significantly associated with mortality in seven days and 30 d (P value < 0.001; Table 5). Table 6 summarises the causes of morality.

Table 4 Correlation of nonsteroidal anti-inflammatory drugs with outcome variables and Forrest classification.

Variables	Total	NSAIDs
		Yes (n = 113)	No (n = 29)
Hospital stay (d)
≤ 5	104 (73.2)	80 (70.8)	24 (82.8)
> 5	38 (26.8)	33 (29.2)	5 (17.2)
Mortality in 7 d
No	127 (89.4)	100 (88.5)	27 (93.1)
Yes	15 (10.6)	13 (11.5)	2 (6.9)
Mortality in 30 d
No	121 (85.2)	95 (84.1)	26 (89.7)
Yes	21 (14.8)	18 (15.9)	3 (10.3)
Re-bleed in 24 h
No	122 (85.9)	98 (86.7)	24 (82.8)
Yes	20 (14.1)	15 (13.3)	5 (17.2)
Re-current bleed in 7 d
No	116 (81.7)	93 (82.3)	23 (79.3)
Yes	26 (18.3)	20 (17.7)	6 (20.7)
Forrest classification
I	18 (12.7)	10 (8.8)	8 (27.6)a
II	21 (14.8)	17 (15.0)	4 (13.8)
III	103 (72.5)	86 (76.1)	17 (58.6)

^aP < 0.05.

n (%) are reported. P value was calculated by Chi-square/ Fisher Exact test. NSAID: Nonsteroidal anti-inflammatory drug.

Table 5 Association of re-bleed with mortality.

Variables	Total	Mortality in 7 d		Mortality in 30 d
		No (n = 127)	Yes (n = 15)	No (n = 121)	Yes (n = 21)
Re-bleed in 24 h
No	122 (85.9)	117 (95.9)	5 (4.1)a	113 (92.6)	9 (7.4)a
Yes	20 (14.1)	10 (50.0)	10 (50.0)	8 (40.0)	12 (60.0)
Re-current bleed in 7 d
No	116 (81.7)	112 (96.6)	4 (3.4)a	111 (95.7)	5 (4.3)a
Yes	26 (18.3)	15 (57.7)	11 (42.3)	10 (38.5)	16 (61.5)

^aP < 0.01.

n (%) are reported. P value was calculated by Fisher Exact test.

Table 6 Causes of mortality.

Cause	n	%
Re-bleed in 24 h	10	47.6
Re-current bleed in 7 d	1	4.76
Septic shock	3	14.2
Cardiac event	2	9.52
Pulmonary embolism	1	4.76
Others	4	19.04

DISCUSSION

PUD is a disease entity that includes gastric and duodenal ulcers. For almost two centuries, this disease has continued to be a major causal factor for hospital admissions and a menace in terms of morbidity and mortality. The current study highlighted the risk factors and outcomes of peptic ulcer bleed in a resource-limited developing country.

The risk of PUB varies with the type, dose and duration of anti-platelet agents, along with or without other NSAIDs[16]. In our study, the use of antiplatelet agents was not associated with the severity of the ulcer in terms of the Forrest classification. This is likely because of the low sample size. However, its use led to statistically significant one-week and one-month mortality. This is probably the first study from this part of the world to report this increased mortality (both early and late) due to antiplatelet medications.

The majority of NSAIDs related to peptic ulcers belonged to the Forrest 3 class. The lack of association between NSAID use and higher Forrest-class ulcers could be multifold. First, we did not examine the individual dose, type, and duration of NSAID use. Secondly, the total number of patients on NSAIDs was not significant enough to reach statistical significance. However, Liu et al[17], in a similar study, showed similar results: only 17% of the patients had Forrest 1 ulcers.

The role of H. Pylori with PUB is varied and controversial. In a recent study, H. pylori infection increased the risk of PUB in patients with NSAIDs, aspirin and non-aspirin antiplatelet agents[18]. However, previous studies have reported more favourable outcomes with H. pylori-related PUB. In a survey by Chason et al[19], patients with H. pylori-related PUB patients had shorter length of hospital stays and lower rebleeding rates. Our data did not show any association of H. pylori with LOS, FC, re-bleed or mortality.

The reported mortality related to PUB varies widely from 4%-15% depending on mortality solely due to PUB or all-cause mortality and shows a significant variation depending on age and comorbidity[20,21]. Most of the local studies elaborating on mortality have reported combined mortality of PUB along with oesophagal variceal bleeding. However, one such study reported an even higher mortality (26.7% for non-variceal bleeding)[22]. In the current study, one-week and one-month mortality stood at approximately 11% and 15%, respectively. A similar study from the United Kingdom showed a mortality of around 8%[23]. However, as most (87%) of our patients had low-risk ulcers (Forrest 2b and beyond), this high mortality likely represents all-cause mortality rather than sole ulcer bleed-related mortality. This likely explanation is further supported by the evidence that nearly two-thirds of the mortality occurred in the first-week post-bleed. However, since mortality does not correlate with the Forrest classification of ulcers, it shows that patients died from multiple factors, among which GIB acts as a catalytic agent rather than as the sole cause. The case would have been settled if we had access to details of each mortality.

Among the causal agents’ anti-platelets were the only ones leading to statistically significant one-week and one-month mortality. This has been highlighted before by Lanas et al[24], that even low-dose aspirin was independently associated with an increased risk of peptic ulcer bleed (OR: 2.4; 95%CI: 1.8–3.3). As seems logical, patients taking antiplatelet medications have concurrent medical conditions such as hypertension and ischemic heart disease, placing them at an additional increased risk of GIB-related mortality.

Our study clearly shows that re-bleed either at 24 h or one week is strongly associated with one-week or one-month mortality. This finding has been documented in multiple other international studies. Branicki et al[25] have shown that the mortality was increased by 17 folds in the case of a re-bleed. They also concluded that co-morbidity adds to the mortality in case of peptic ulcer bleed. However, this is the first study from this part of the world covering this aspect of peptic ulcer bleed. As mentioned earlier, this mortality rate is assumed not to be solely attributable to peptic ulcer bleed or re-bleed but a combination of bleed with the comorbid conditions.

Our study shows that alternative (Herbal) medications influence 24-hour or 7-day re-bleed risk. A local research from China in 2016 showed that around 7% of the patients presenting with upper GI bleed used Chinese herbal medicines[26]. The content and composition of these medications are variable and, in most cases, are unknown. At times, they contain a mixture of NSAIDs along with Steroids to treat a variety of medical conditions and musculoskeletal pain. Tomlinson et al[27] documented in their study that herbal medicines contaminated with NSAIDs or steroids can lead to an increased risk of peptic ulcer bleed or even perforation. Our study also shows that a sizeable number (40%) of patients were using these alternative medications. This reflects a lack of education and less access to proper healthcare facilities.

The risk of re-bleeding from FC 1 ulcer in our study (39%) is close to the internationally reported risk of around 50%[28]. However, none of the risk factors was associated with the Forrest classification of ulcers. Moreover, the Forrest classification did not correlate with outcome variables like LOS, Re-bleed or Mortality. This could be due to a small number of patients (12.7%) having Forrest 1 type ulcers. Giese et al[29] also showed that the Forrest classification did not correlate with these end-point variables as ours. In our study, one week's re-bleed rate from Forrest 3 ulcers (13.6%) is more than the reported initially (0%-10%)[30]. Other factors, such as medications, could have influenced this; however, given the small number of patients, we were unable to perform any multivariate analysis.

In the current study, nearly three-quarters of patients were discharged within five days of admission. This is in line with international data. Schacher et al[31], in a study from Switzerland aimed at the length of hospital stay and timing of endoscopy, showed a similar length of hospital stay of around five days, which did not show a statistical difference among patients undergoing early or late endoscopy. A much larger cohort of patients from the United Kingdom that investigated the LOS among a similar group of patients showed a median LOS of around six days[23]. None of the etiological variables or the Forrest classification correlated with LOS. The likely explanation for this high number of early discharges in our study is the low number of patients with high-risk Forrest ulcers (1 and 2a).

In this study, PUD affected patients of all ages. Though subgroup analysis of age groups could not be done due to a low number of patients, this goes hand in hand with the local studies[15].

While this study breaks new ground in analysing PUB risk factors and outcomes for a Pakistani population, its impact is tempered by some limitations. Its pioneering nature and focus on clinically relevant factors such as antiplatelet use and comorbid conditions offer valuable information for local healthcare providers. However, the small sample size and single-centre design restrict its generalizability and necessitate further research with more extensive, diverse samples. Additionally, the lack of data on specific details of medications hindered a more precise understanding of their influence on outcomes. Moving forward, larger studies and targeted investigations into medication profiles are crucial to solidify the findings and personalise treatment protocols. Ultimately, incorporating this research alongside future studies has the potential to inform tailored clinical practice guidelines for Pakistani PUB patients, leading to significantly improved care.

CONCLUSION

Antiplatelet use and rebleeding increase the risk of early mortality in PUD-related UGIB, while alternative medicines are associated with early rebleeding.