A total of 254 cases of severe preeclampsia were received and managed at the YGOPH, from 1 January 2020 to 31 July 2021. Seventeen of them presented with HELLP syndrome (frequency: 6.7%), amongst which one patient who presented with hepatitis B was excluded; 16 patients were retained. The mean age was 27.19 ± 6.44 years, while the most represented age group in our study were those between 20 and 24 years (31.3%) (Figure 1). Of these participants, 41.7% were employed, 50% were unemployed, and 8.3% were students. In addition, 33.3% were single and 66.7% were married, while 50% had only primary education, and those that had no education, secondary and higher level of education each consisted of 16.7% of the study population.

The 16 cases retained were divided into: 6 (37.5%) class 1, 9 (56.2%) class 2, and 1 (6.3%) class 3, according to the Mississippi’s classification of HELLP syndrome [11] . Following the Mississippi classification, 37.5% of our study participants had severe HELLP syndrome, 56.2% had moderate HELLP syndrome,

6.3% participants had mild HELLP syndrome.

In the course of treatment, 9 patients died while 7 survived. The majority of our study participants have had no delivery (33.3%), alongside those who have had two deliveries at term (33.3%). Similarly, the majority, 56.3% of our study participants had antenatal contacts of poor quality, while three (18.8%) of the participants had no contacts. Only two (12.5%) participants had good quality antenatal contacts, according to WHO’s recommendations. All the study participants who were referred, did their antenatal contacts in the different category hospitals in Cameroon, including specialized clinics, with the majority coming from integrated health centers and specialized clinics respectively (31.3% each).

Concerning preeclampsia, the diagnosis was missed during antenatal contacts in up to 25% of cases, meanwhile the diagnosis was made in 18.8% of our study participants, but 12.5% of their health providers did not put on treatment and closely follow-up their patients, while 6.3% initiated antihypertensive treatment. The rest were revealed either as HELLP syndrome or eclampsia, which are all complications of preeclampsia.

Concerning HELLP Syndrome, the majority of cases were diagnosed during post-partum (56.3%), three cases (18.8%) were diagnosed during labor, while the rest were either diagnosed during antenatal contacts or brought in emergency to the hospital.

Contrary to the literature, where HELLP syndrome usually manifests within the first 48 hours post-partum, two cases (22.2%), were diagnosed beyond 48 hours (Figure 2).

As seen in Table 1 below, our study participants manifested common clinical features, amongst which were: elevated blood pressure, headache, epigastric pain, convulsions, nausea and dizziness, which were remarkable at: 100%, 93.8%,

87.5%, 75.0%, 68.8%, and 62.5% respectively.

Concerning delivery, 62.5% (10) of our study participants delivered by cesarean section, while 37.5% (6) delivered vaginally. Though all referred cases, 56.2% delivered at the YGOPH, while 43.8% delivered at the health centers where follow-up was done. The majority of our study participants received Magnesium sulphate (81.2%) and Nicardipine (87.5%). However, few participants received Alpha Methyl Dopa (31.2%) and Labetalol (18.8%).Thirteen (81.2%) of our study participants received anti-convulsant; diazepam and phenobarbital, while 4 each (25.0%) further received Midazolam and Fentanyl (Table 2).

All our study participants were rehydrated with crystalloids; the mean volume of crystalloids administered was 2938 ± 310 ml. Concerning transfusions, 60%

*SD: Standard deviation.

(9) participants received whole blood, 31.2% (5) received fresh frozen plasma, while 46.7% (7) received platelets concentrate. Of the 16 participants with HELLP syndrome, 50.0% (8) received the LMWH enoxaparin (Lovenox®) during their management, while only 4 out of 11 who delivered preterm received corticosteroids, with the aim of maturing fetal lungs.

As seen in Table 3 below, over 50% of our study participants presented severe complications after delivery, some of which were suspected without paraclinical confirmation, albeit with strong clinical suspicion. Regarding the type of complications, a majority had altered consciousness (87.5%), followed by disseminated intravascular coagulation (68.8%), and suspected stroke and acute pulmonary edema at 56.2% each. It is important to note that stroke was confirmed in 12.5% of our participants. However, 43.8% (7) cases recovered with blood pressure normalization occurring in 6 (37.5%) participants, while 56.2% (9) finally died.

Table 4 below presents the gestational ages of all participants at delivery, foetal birthweights, and the Apgar score at fifth minute.

Table 4 below shows the gestational ages, birthweights and the Apgar score at the 5^th minute of the neonates delivered during the study period. Birthweights were missing for 4 neonates. So, we could not ascertain whether there was low birth weight or IUGR.

Concerning fetal outcome assessment, the majority, 68.7% (11) neonates were delivered preterm, while 31.3% (5) were delivered at term. Amongst them, 31.2% (5) were still births, 12.5% (2) had intrauterine growth restriction or low birth weight. Three (18.8%) of the neonates presented with a poor Apgar score at the 5th minute, while 50% finally had a good Apgar score. Placenta abruptio was identified in one case of still birth (Table 5).

Compared to those who survived (43.8%), participants who died (56.2%) presented more severe complications (suspected stroke, P < 0.001; suspected hepatic hematoma/rupture, P = 0.001; disseminated intravascular coagulation, P = 0.005;

*Suspected: Strong clinical suspicion in patient’s file, but without paraclinical confirmation. **Disseminated intravascular coagulation: based on excessive bleeding, ecchymosis and/or petechia.

*Multiple pregnancy.

*Still births included one term and four pre-term deliveries; **Based on birthweights adjusted for gestational age; LBW: Low birth weight.

*Suspected: Strong clinical suspicion in patient’s file, but without paraclinical confirmation; **Disseminated intravascular coagulation: based on excessive bleeding, ecchymosis and/or petechia; ^aChi squared test or Fischer exact test (if counts < 5).

and normalized blood pressure, P = 0.001). Other clinical features like altered consciousness, confirmed stroke, suspected pulmonary edema, and acute kidney injury were also higher in those who died, albeit without statistical significance. The blood pressure took between 11 to 35 days to normalize in survivors. However, there was no blood pressure normalization in those who died (Table 6).

^aChi squared test or Fischer exact test (if counts < 5); ^bMann-Whitney U test; SD: Standard deviation.

^aChi squared test or Fischer (if counts < 5); *Patient was referred to another hospital for dialysis; NA: Not applicable.

More participants who died received fluid resuscitation and blood products than those who survived, but the difference was not statistically significant. All participants were hydrated with crystalloids, and over 50% in each subgroup (survived vs died) received whole blood transfusions. A third of the participants in each group received fresh frozen plasma, while over 50% of those who died received an additional platelet transfusion. Although not statistically significant, those who died received more crystalloid resuscitation volume whereas those who survived received a greater volume of whole blood transfusions on average (Table 7).

As seen in Table 8 above, none of the participants who died did any of the investigations mentioned. Of the survivors, one participant did an abdominal ultrasound, one did a CT scan, which revealed ischemic stroke and she was adequately managed, while 1 participant with AKI was finally sent for dialysis.

The frequency of HELLP syndrome in our setting (6.7%) is different from that stated in the literature (10% - 20%) [11] . African statistics are fairly incomplete regarding the incidence of this syndrome. However, Noubiap et al. in their meta-analysis of hypertensive disorders in pregnancy in the African setting found that HELLP syndrome occurred in 2.2% of patients [5] . This prevalence could have been lower if it was assessed amongst only patients with severe preeclampsia as in the present study. In Cameroon, Essiben et al. in their study of eclampsia, found that 9.9% of patients with eclampsia developed HELLP syndrome [8] . This result is higher than ours probably because unlike the present study, they studied patients with eclampsia, which are a subset of the patients with severe preeclampsia. Another Cameroonian retrospective study by Kemfang et al. in 2015 [12] showed that 1.8% of patients with severe preeclampsia developed HELLP syndrome. Their findings may have differed from ours because their study involved 3 referral hospitals with a larger sample size unlike our single-centre study. The values in our setting are lower than those in the literature probably due to under-diagnosis and the fact that most women visit health centres for delivery.

We found that the most represented age group was those between 20 and 24 years (31.3%), while the least represented, were those between 15 to 19 years (12.5%) and 35 to 39 years (12.5%) respectively. However, the majority of patients (56.3%) were aged 25 years and above. This corroborates with the literature, where HELLP has been shown to occur in older maternal age groups, with a mean age of 25 years [11] [13] . Essiben et al. equally had the 20 to 25 year age group as the most represented at 31.8% in patients with eclampsia [8] . Age is a well-established risk factor of complications during pregnancy, including hypertensive disorders in pregnancy [5] .

Most patients (56.3%) had poor follow up and were gravida one and two respectively. Moreover, most of our study participants were nulliparous (33.3%) or have had two deliveries at term (33.3%). This differs from the prevalence of HELLP, which is more frequent amongst multiparous women stated in the literature [13] . However, our findings are similar to those of Essiben et al. and Fouedjio et al. who studied eclampsia in Cameroon [7] [8] . Tebeu et al. equally found that older women had fewer than four antenatal contacts in the north region of Cameroon [14] . This could be because in our setting, most women visit specialized health care facilities during their first delivery, and decrease this tendency with increasing parity. In effect, a systematic review of antenatal consultations in developing countries found a statistically significant negative effect of parity on the adequacy of antenatal consultation [15] . In unspecialized health centres, there is an increased tendency of missed diagnosis as these lack the appropriate diagnostic equipment and trained personnel required for the prompt detection of this fatal disease. The mortality rate can be reduced by regular antenatal contacts, good follow-up and transfer of pregnant women who present with the risk of developing HELLP syndrome to the intensive care unit without delay.

In the present study, the majority of cases of HELLP syndrome were diagnosed during post-partum (56.3%), three cases (18.8%) were diagnosed during labour, while the rest were brought to the hospital in emergency. Conversely, Sibai et al. [16] found that 70% of HELLP syndrome cases occur before delivery. The difference may be attributed to the fact that most of the patients visit health centres that lack specialists and adequate diagnostic tools. Those who eventually go to tertiary hospitals are either referred late from the aforementioned health centres or arrived, there is misdiagnosis or their results are made available late due to financial constraints or bureaucracy. Concerning preeclampsia, the diagnosis was missed during antenatal contacts in up to 25% of cases, meanwhile the diagnosis was made in 18.8% of our study participants, but 12.5% of their health providers neglected the treatment and close follow-up, while 6.3% initiated antihypertensive treatment. The rest were revealed either as HELLP syndrome or eclampsia, which are all complications of preeclampsia. None of the patients who died in the present study underwent abdominal ultrasound, chest x-ray, head CT scan, or dialysis care. This mirrors the underdiagnosis due to inappropriate treatment centres.

Our study participants manifested common clinical features, amongst which where: headache, epigastric pain, convulsions, nausea and dizziness, which were remarkable at: 93.8%, 87.5%, 75.0%, 68.8%, and 66.7% respectively. These corroborates with findings from the literature [17] [18] ; Sibai et al. and Aarnoudse et al. reported similar findings in their studies. The only difference between their studies and ours was that epigastric pain was the most common symptom unlike headache in the present study. This may due to the fact that epigastric pain is a common neglected symptom in our setting, which may hide a serious underlying pathology. Physicians should therefore pay attention to this seemingly trivial symptom, especially in pregnant women. Women who present with headache or epigastric pain during contacts should be thoroughly investigated in order to identify the underlying cause and treat accordingly.

Ten (62.5%) of our study participants delivered by cesarean section, while 6 (37.5%) delivered vaginally. All 6 per-vaginal deliveries had complications; hence, it may be better to proceed to the delivery by caesarien section that is recommended in the literature [13] [19] . Sibai et al. recommended that a trial of labor be done only in patients with mild to moderate HELLP syndrome who are stable, have a favorable cervix and are at 32 weeks of gestation or greater [19] .

The majority of our study participants received Magnesium sulphate (81.2%) and Nicardipine (87.5%). However, few participants received Alpha Methyl Dopa and Labetalol. This is similar to the findings of the single-centre study of Ahmad et al. in Egypt; all patients received Magnesium sulphate and most received antihypertensives (nifedipine; 62%, labetalol; 9.2%) [20] . In our study, no patient who died had a normalisation of blood pressure, whereas 6 (85.7%) of the 7 patients who had attained blood pressure normalisation (P = 0.001). This could imply that appropriate antihypertensive treatment could be life-saving in these patient group. However, larger multi-centre studies are required to confirm this finding.

Thirteen (81.2%) of our study participants received anti-convulsant; diazepam and phenobarbital, while 4 (25.0%) further received Midazolam and were also sedated with Fentanyl. Sedatives were mostly given to patients who were comatose, and all eventually died.

The treatment of HELLP syndrome is symptomatic. However, in our setting, we do not always have readily available blood products like fresh frozen plasma and platelets, and most patients also get them late or not at all due to financial constraints. All our study participants were resuscitated with crystalloids; the mean volume of crystalloids administered was 2938 ± 310 ml. Equally, 9 (60%) participants received whole blood, 5 (31.2%) received fresh frozen plasma, while 7 (46.7%) received platelets concentrate. On average, patients received 4000 ml of fluids per day, which may have predisposed to fluid overload. Compared to patients who survived, those who died received more crystalloids and blood, though the blood products compared to the severity of the disease may not have been sufficient or adapted. A study by Jashmid et al. showed that plasmapheresis is a better life-saving treatment in HELLP syndrome patients compared to other blood products [21] .

In the present study, corticosteroids (first dose) were administered in only 4 participants, with the aim to mature the foetal lungs. Routine administration of corticosteroids in patients with HELLP syndrome is recommended. The antenatal administration of dexamethasone at a high dosage of 10 mg intravenously every 12 hours has been shown to markedly improve the laboratory abnormalities associated with HELLP syndrome [22] . Patients treated with dexamethasone exhibit longer time to delivery [13] ; this facilitates maternal transfer to a tertiary care centre and postnatal maturity of foetal lungs. Therefore, appropriate corticosteroid administration may reduce materno-foetal mortality.

Concerning outcome assessment, over 50% of our study participants presented severe complications after delivery, ranging from: eclampsia, sepsis, stroke, hepatic rupture, acute pulmonary edema, disseminated intravascular coagulation, and acute kidney injury. Some of these complications were suspected without paraclinical confirmation. However, 43.8% recovered, while 56.2% finally died. This may be due to the fact that there was misdiagnosis, delayed diagnosis, financial constraints, and unadapted treatments. Isler et al. found cerebral haemorrhage or stroke to be the primary cause of death in 26% and the most contributing factor in another 45% of the deaths [23] . Dan et al. in their study found that maternal and perinatal mortality rate in hepatic rupture ranges from 18 to 86% and 80%, respectively [24] . The onset of acute kidney injury has been associated to increased maternal mortality in HELLP syndrome [25] . In another study, disseminated intravascular coagulation (DIC), sepsis/multi-organ dysfunction, and massive pulmonary embolism were found to be causes of death [26] . In the present study, all comatose patients died, and suspicion of stroke (P < 0.001), hepatic rupture (P = 0.001), suspected acute pulmonary edema (P = 0.615), disseminated intravascular coagulation (P = 0.005), and acute kidney injury (P = 0.103) were higher in patients who died than in those who survived. Unfortunately, we cannot with certainty say the exact cause of death since the necessary investigations to confirm the diagnosis were not done. However, they were suspected as being possible causes of death. The proper diagnosis of organ dysfunction in patients with HELLP syndrome could enable us to properly manage them and prevent adverse materno-fetal outcomes. In a large retrospective cohort study comprising 442 pregnancies complicated by the HELLP syndrome, maternal mortality was 1.1% [16] , which is in accordance with other reports [17] [27] . However, there have been reports of higher maternal mortality, up to 25% [28] . Ideally, unexpected rapid death from HELLP may require forensic expertise and further investigation [29] . The deaths of the 9 (56.3%) mothers and the 5 (31.3%) neonates in the present study could have been prevented if these investigations were done on time to identify the possible complications and give adequate treatment.

Regarding the foetal outcome assessment, the majority (68.7%) neonates were delivered preterm, while (31.3%) were delivered at term. Amongst them, 5 (31.2%) were still births, 2 (12.5%) had intrauterine growth restriction or low birthweight. Eighteen (18) percent of the study participants had a persistent poor Apgar score at the 5th minute, while 50% finally had a good Apgar score. As per the literature, the perinatal mortality rate related to the HELLP syndrome is between 7.4% and 34% [17] [30] , similar to that in the present study. Neonates delivered before 32 weeks’ gestation have the highest risk of perinatal death. According to Gul et al., the perinatal mortality was 34% before 32 weeks’ gestation, and 8% after the 32nd gestational week [30] . In the present study, 2 (13.3%) of the 5 neonates delivered before 32 weeks’ gestation were still births, alongside 2 (13.3%) delivered after the 32nd week died, and one neonate delivered at term. This differences from the reported data in the literature reflect the misdiagnosis and neglected diagnosis in our setting. Many neonates may have died before 32 weeks’ gestation, but may have passed away as still births due to other causes, without the diagnosis of HELLP or preeclampsia ever being made. The lower frequency of HELLP in our context testifies to this. Prematurity, placental insufficiency, with or without intrauterine growth restriction and abruptio placentae, are the leading causes of neonatal death [11] . Early diagnosis and the prompt initiation of appropriate treatment will definitely improve foetal outcome.