LDL Cholesterol Goal Attainment and Statin Intensity Profile Among Patients with Atherosclerotic Cardiovascular Disease: A Real-world Study in Private Hospitals in Indonesia

Cardiovascular disease (CVD) continues to be the foremost cause of mortality worldwide, accounting for an estimated 19.8 million deaths in 2022 – approximately one-third of all global deaths.¹ Indonesia faces a similar burden: in 2019, CVD was responsible for 38% of all fatalities, with stroke and ischaemic heart disease, the leading clinical forms of atherosclerotic CVD (ASCVD), ranking as the country’s top two killers, contributing around 19 and 14% of deaths, respectively.² These figures underscore the immense health burden of ASCVD and the urgency of improving secondary prevention efforts in high-risk populations. Effective management of established ASCVD is a public health priority to curb preventable morbidity and mortality.

Dyslipidaemia, particularly elevated LDL cholesterol (LDL-C), is a major modifiable risk factor in the pathogenesis and progression of ASCVD.^3,4 Robust evidence from meta-analyses of large randomised controlled trials has demonstrated that lowering LDL-C yields proportional reductions in cardiovascular events – approximately a 20% relative risk reduction in major vascular events per 1.0 mmol/l decrease in LDL-C levels.^5,6 Accordingly, current clinical guidelines advocate aggressive lipid-lowering therapy for secondary prevention. For very-high-risk patients, such as those with established ASCVD, an LDL-C goal of ≤1.4 mmol/l (with at least a 50% reduction from baseline) is now recommended to maximise risk reduction.^7,8 Achieving such stringent LDL-C targets, however, can be challenging in routine practice.

Despite the proven benefits of statin therapy, real-world data reveal persistently low rates of LDL-C goal attainment among ASCVD patients. Multiple contemporary studies in the Asian region have reported suboptimal lipid control in secondary prevention cohorts.^9,10 For example, an international survey (CEPHEUS) found that only about half of treated patients achieved their LDL-C goal overall, and in Indonesia, this proportion was even lower (≈31% LDL-C goal attainment). Notably, among very-high-risk Indonesian patients requiring an LDL-C <1.8 mmol/l, only ~12% reached that target under standard care.^11,12 Such findings highlight a substantial care gap, where the majority of ASCVD patients fail to attain guideline-recommended LDL-C levels despite being on therapy. Key contributors to this gap include therapeutic inertia and underutilisation of intensive statin regimens.

In many Asian settings, there appears to be reluctance among clinicians to prescribe high dose or high-intensity statins for secondary prevention, due to concerns about side-effects or other barriers.^13–15 This suboptimal use of potent lipid-lowering therapy, combined with issues of adherence and access, has resulted in LDL-C being one of the least adequately controlled risk factors in secondary prevention.

In Indonesia, most existing research on dyslipidaemia management and outcomes has focused on patients in the public health system or broader population-based samples, whereas data from the private healthcare sector remain limited. This gap is important, because private hospitals cater to a substantial segment of ASCVD patients, as approximately 52% of Indonesian hospitals are privately operated, and treatment patterns in these settings may differ from the public sector.² Notably, until recently, Indonesia’s national health insurance covered only low- to moderate-intensity statins – not high-intensity therapy – which may have constrained the use of more intensive regimens in public practice.⁹ It is therefore unclear whether patients in private hospitals, who might have access to different medications, managed to achieve better lipid control. In light of the overall low LDL-C target attainment, and the potential discrepancies in care between private and public hospitals, we designed the present study to characterise LDL-C goal attainment and lipid-lowering treatment patterns among Indonesian patients with clinical ASCVD in a real-world private hospital setting. In particular, we aimed to assess the proportion of patients achieving the recommended LDL-C goal of ≤1.4 mmol/l and to identify predictors of LDL-C goal attainment, with special attention to the impact of statin intensity (moderate- versus high-intensity therapy) on lipid outcomes. Our findings will help delineate current gaps in secondary prevention of ASCVD events, and inform whether our current clinical management is sufficient to achieve optimum LDL-C outcomes in real world settings.

Methods

Study Design and Setting

We conducted a multicentre, retrospective cross-sectional study using routinely collected electronic medical records (EMR) from two private hospitals in Greater Jakarta, Indonesia (Siloam Hospitals Kebon Jeruk; Siloam Hospitals Lippo Village). Both sites are private hospitals in the Siloam Hospitals group, and provide inpatient and outpatient cardiovascular care. The hospitals were selected a priori based on location in the Greater Jakarta area, to maximise diversity of the patient pool and comparability with the national population. The EMR was queried for encounters between 1 January 2022 and 31 October 2023. For each eligible patient, a single index encounter within this window was defined a priori, and all variables (including LDL-C and statin intensity) were analysed as point-in-time measures at that index. Reporting follows STROBE guidance for cross-sectional studies.

Study Population and Sample Selection

Inclusion criteria were adults (aged ≥18 years) with established clinical ASCVD, identified by ICD-10 codes and/or procedure codes for any of the following: prior acute MI, acute coronary syndrome/unstable angina, ischaemic stroke or transient ischaemic attack, aortic aneurysm, peripheral arterial disease, or a history of coronary or other arterial revascularisation. Patients were required to be on current statin therapy (any intensity) at the index encounter. We imposed no exclusions by comorbidity profile or admission status (inpatient/outpatient). A categorical descriptive formula was used to determine minimum sample size in this study, as the following:

n=Zα² P·Q / d²

with Zα=1.96 and d=0.05 (1)

Using the proportion from a previous study, 417 (379 + 10%) patients are used to estimate the population.¹⁶

Data Collection

All clinical variables, medication orders and laboratory values were extracted from the EMR by trained hospital data teams using uniform field definitions. Harmonisation rules prespecified how to reconcile variables that could be captured in multiple fields (e.g. diagnosis list versus procedure history). Quality checks addressed duplicate rows, non-patient artefacts (e.g. header/footer lines), out-of-range values and logical inconsistencies (e.g. invalid sex codes); discrepancies were queried and resolved before data lock. Analyses were prespecified as complete case with respect to the variables used in each model; no statistical imputation was planned. The data collection diagram is presented in Figure 1.

Variables and Operational Definitions

The primary outcome was the attainment of LDL-C ≤1.4 mmol/l at the index encounter, defined a priori to reflect contemporary secondary prevention guidance for established ASCVD (2024 European Society of Cardiology Guidelines for the Management of Chronic Coronary Syndromes; 2023 American College of Cardiology and American Heart Association Guideline for the Management of Patients with Chronic Coronary Disease).^7,8 We did not separately evaluate more stringent ‘extreme-risk’ targets (e.g. <1 mmol/l), because a number of patients were not on maximally tolerated statins at the time of an ASCVD event. Classifying patients with a history of multiple ASCVD events into the ‘extreme-risk’ category without consistent data regarding consumption of maximally tolerated statin will lead to bias in analysis. The main exposure for comparative effectiveness was statin intensity at index, categorised as low, moderate or high based on EMR records and standardised accordingly.

For the intensity analysis, only moderate- and high-intensity statin users were compared, while patients on low-intensity therapy (if any) were excluded. Covariates, selected a priori and treated as binary (1/0), included age ≥60 years, female sex, hypertension, diabetes, chronic kidney disease, history of ASCVD (composite of relevant conditions) and length of statin therapy (LoT) ≥180 days prior to the index LDL-C measurement. Both age and LoT were modelled categorically by design, to align with local decision-making practices and to minimise extrapolation from potential non-linear continuous effects.

Bias and Confounding Control

For the moderate- versus high-statin comparison, we anticipated confounding by indication. We therefore estimated inverse probability of treatment weights (IPTW) based on a logistic propensity score for receiving high-intensity therapy conditional on the prespecified covariates (age ≥60 years, female sex, hypertension, diabetes, chronic kidney disease, ASCVD history, LoT ≥180 days). Stabilised weights were used and trimmed at the 1st to 99th percentiles to limit the impact of extreme weights and improve precision.¹⁷ We quantified covariate balance with absolute standardised mean differences, using standardised mean differences <0.10 as indicative of adequate balance.¹⁸ Misclassification risk was minimised by using coded EMR fields and consistent operational definitions across sites.

Data Analysis

All analyses were conducted as two-sided tests with a significance threshold of α=0.05. Categorical variables are presented as counts and percentages, while continuous variables are summarised as mean ± SD or median with interquartile range, as appropriate. Analyses were performed in R version 4.5.1 (R Foundation for Statistical Computing) using the packages logistf, sandwich, lmtest, tidyverse, pROC and ResourceSelection.

To evaluate predictors of achieving LDL-C ≤1.4 mmol/l, we prespecified Firth’s bias-reduced logistic regression to address potential small-sample bias and separation, as a low prevalence of this outcome was expected in real-world ASCVD study. The model incorporated seven covariates (age ≥60 years, female sex, hypertension, diabetes, chronic kidney disease, ASCVD history and length of statin therapy ≥180 days before the index LDL-C measurement), with results reported as adjusted ORs and 95% CIs. Model performance was assessed through discrimination (area under the receiver operating characteristic curve) and calibration (Hosmer–Lemeshow test), while acknowledging the limitations of these diagnostics in modest sample sizes.

For the comparison of moderate- versus high-statin intensity, we applied IPTW based on a propensity score model fit with logistic regression (base R) to estimate population average treatment effects. Weighted Poisson regression with a log link and robust (sandwich) variance was used to obtain RR, and a weighted linear probability model with robust variance was used to estimate risk differences. Covariate balance before and after weighting was evaluated through standardised mean differences (base R) and visual inspection of propensity score overlap (plots generated with ggplot2). As a sensitivity analysis, we also fit a covariate-adjusted Firth’s logistic model including statin intensity and the same covariates to verify consistency of effect direction and magnitude with the IPTW estimand.

Ethical Considerations

The protocol was approved by the Mochtar Riady Institute for Nanotechnology Ethics Committee (approval 022/MRIN-EC/ECL/IX/2023, dated 11 September 2023), and conducted in accordance with the Declaration of Helsinki. EMR data were de-identified prior to analysis; consent procedures followed the approved protocol and institutional policies.

Use of Generative Artificial Intelligence

Generative AI (ChatGPT-5 Pro, OpenAI) was used exclusively to enhance grammar and sentence clarity. All intellectual content, analysis and conclusions were developed by the authors, and every output from the tool was critically reviewed and approved prior to inclusion in the manuscript.

Results

Baseline Characteristics

We analysed 422 individual patients drawn from two private hospitals in the Greater Jakarta region. All variables used in the analyses were complete. Men comprised 312 of 422 patients (73.9%), and women comprised 110 of 422 patients (26.1%; Table 1 ). Age distribution was <40 years 3.1%, 40–49 years 12.3%, 50–59 years 31.0%, 60–70 years 32.0% and >70 years 21.6%; age ≥60 years was present in 226 of 422 patients (53.6%). Comorbidities included hypertension 236 of 422 patients (55.9%), diabetes 154 of 422 patients (36.5%) and chronic kidney disease 42 of 422 patients (10.0%). A history of ASCVD events prior to the recruitment point (January 2022 to October 2023) was documented in 235 of 422 patients (55.7%). Among ASCVD phenotypes, previous revascularisation was recorded in 176 of 422 patients (41.7%), cerebral infarction in 162of 422 patients (38.4%), acute MI in 108 of 422 patients (25.6%), unstable angina in 41of 422 patients (9.7%), transient ischaemic attack in 26 of 422 patients (6.2%), peripheral arterial disease in 17 of 422 patients (4.0%) and aortic aneurysm in eight of 422 patients (1.9%; patients could have multiple conditions) (Table 1). At presentation, LoT ≥180 days was present for 204 of 422 patients (48.1%). Admission type was outpatient for 109 of 422 patients (25.8%), and inpatient for 313 of 422 patients (74.2%); among inpatients, length of stay had a median of 5 days (interquartile range 2–8 days; range 1–34 days) (Table 2 ).

LDL Cholesterol Profile and Statin Intensity

At the index assessment, statin intensity was moderate in 279 of 422 patients (66.1%), high in 142 of 422 patients (33.7%) and low in one of 422 patients (0.2%; Table 3 ). The overall mean LDL-C was 2.3 ± 0.9 mmol/l. LDL-C ≤1.4 mmol/l was achieved in 51 of 422 patients (12.1%).

Predictors of Achieving LDL Cholesterol ≤1.4 mmol/l

A prespecified Firth’s penalised logistic regression including seven covariates was fit in 422 patients with 51 events, as shown in Figure 2. Adjusted ORs with 95% CIs were as follows: hypertension OR 0.47 (95% CI [0.25–0.87]), p=0.017; length of statin therapy ≥180 days OR 0.57 (95% CI [0.30–1.05]), p=0.073; chronic kidney disease OR 0.68 (95% CI [0.18–1.92]), p=0.502; female sex (versus male) OR 1.05 (95% CI [0.51–2.05]), p=0.899; prior ASCVD OR 1.06 (95% CI [0.56–2.03]), p=0.870; age ≥60 years OR 1.15 (95% CI [0.62–2.15]), p=0.648; and diabetes OR 1.17 (95% CI [0.62–2.20]), p=0.624.

Moderate- versus High-intensity Statin and LDL Cholesterol Goal Attainment

For the intensity comparison, the single low-intensity case was excluded, leaving 421 patients (moderate n=279, high n=142). Unadjusted event rates showed that LDL-C ≤1.4 mmol/l was achieved in 25 of 279 patients (9.0%) on moderate-intensity statins, and 26 of 142 patients (18.3%) on high-intensity statins (Figure 3 ). The corresponding unadjusted RR for high versus moderate intensity was 2.04 (95% CI [1.23–3.40]), with a risk difference of +9.3 percentage points (95% CI [+2.2, +16.5]). In the IPTW analysis, propensity scores for high-intensity therapy were estimated from age ≥60 years, female sex, hypertension, diabetes, chronic kidney disease, prior ASCVD and length of statin therapy ≥180 days, with stabilised weights trimmed from the 1st to the 99th percentiles. After weighting, the marginal RR was 2.23 (95% CI [1.30–3.80]) and the marginal risk difference was +10.7 percentage points (95% CI [+3.0, +18.4]). Covariate balance improved across all measured variables, with weighted absolute standardised mean differences <0.10 (Supplementary Figure 1 ), and propensity-score distributions demonstrated adequate overlap between treatment groups (Supplementary Figure 2). Sensitivity analysis using a covariate-adjusted Firth’s logistic model with the same covariates yielded an adjusted OR of 2.53 (95% CI [1.37–4.72]; p=0.003).

Discussion

In this real-world, cross-sectional study from two private hospitals in Greater Jakarta, attainment of the recommended LDL-C target in ASCVD patients (≤1.4 mmol/l for very-high-risk secondary prevention) was uncommon (12.1%). Such suboptimal control of LDL-C is concerning, because elevated LDL-C is a well-established driver of atherosclerosis and recurrent cardiovascular events.^4,19 Each ~1 mmol/l reduction in LDL-C has been shown to reduce major vascular events by ~20%.^20,21 Failure to reach LDL-C goals thus likely translates into a substantially higher residual risk of MI, stroke and other adverse outcomes in our population.

Evidently, multiple factors may contribute to this failure. For instance, our study shows that only 33.7% of patients were prescribed high-intensity statins, while the majority of patients (66.1%) were prescribed moderate-intensity statins. Of patients being prescribed high-intensity statins, 57.7% were given atorvastatin 40 mg, while the rest were prescribed rosuvastatin 20 mg. Similarly, a study conducted in the US reported 60.6% of ASCVD patients were not prescribed high-intensity statin.²² Sarraju et al. also revealed that after 1 year upon experiencing an ASCVD event(s), less than one-quarter of Californian patients were given high-intensity statins, with nearly half of their patients not being prescribed any statin.²³ Considering current guidelines recommend prescribing maximally tolerated high-intensity statins as first-line therapy in very-high-risk patients, this phenomenon suggests that therapeutic inertia may be present worldwide, namely, due to low clinician awareness of when to intensify treatment dosage, low medication compliance and so on.^7,8

Furthermore, our findings indicate high-intensity statin therapy was markedly more effective than moderate-intensity therapy for goal attainment, with an IPTW-adjusted RR of 2.23 (OR 2.53). This aligns with the expected efficacy of high-intensity statins according to current guidelines, which typically lower LDL-C by ≥50% versus ~30–49% with moderate-intensity statins. It also mirrors prior evidence that intensive lipid lowering confers additional clinical benefits in high-risk patients.²⁴

However, it is notable that even with high-intensity statins, a majority of our patients did not attain LDL-C ≤1.4 mmol/l, highlighting that monotherapy may be insufficient for many very-high-risk individuals. In a recent Spanish study of patients on high-intensity statins, 71.7% of very-high-risk patients still failed to reach their LDL-C goals, consistent with our observation, and suggesting that additional intensification (such as ezetimibe or proprotein convertase subtilisin/kexin type 9 [PCSK9] inhibitors) may be required in a substantial proportion of cases.¹⁹ DA VINCI, a multinational European study, further corroborates this, as only approximately one-fifth of very-high-risk patients consuming high-intensity statin were reported to achieve their target LDL-C concentration of <1.8 mmol/l.²⁵ For secondary prevention, contemporary guidelines now recommend maximally tolerated high-intensity statins, with add-on therapy as required to reach ≥50% LDL reduction, and <1.4 mmol/l in very-high-risk patients.^7,8

When comparing our results with other contemporary studies, it becomes clear that low LDL-C target attainment is a pervasive problem, especially under newer goal thresholds. Our 12.1% attainment rate is lower than earlier reports with less stringent goals, but parallels findings from Asian and global cohorts in recent years. For instance, the multicountry DYSIS II registry (2013–2014) reported that only ~31% of stable coronary patients in Asia-Pacific achieved LDL-C <1.8 mmol/l despite >90% of patients being on statin therapy. Among post-acute coronary syndrome patients in that registry, just 23% had LDL <1.8 mmol/l at hospital admission, improving to ~42% at 4 months post-discharge with intensified treatment. Notably, mean statin doses in DYSIS were modest (atorvastatin-equivalent ~20 mg/day) and use of adjunctive ezetimibe was low (≈13% of patients). The investigators concluded that while lipid-lowering therapy was common, it was ‘not used to its full potential’ in real-world Asian practice.¹³

These observations resonate with our study – undertreatment (whether due to submaximal dosing, suboptimal statin selection or lack of combination therapy) emerges as a key theme underlying poor goal attainment. More recent data with the <1.4 mmol/l target reinforce this concern. In a study from India, for example, only 20.9% of acute coronary syndrome patients managed to reach LDL-C <1.4 mmol/l after 1 year on high-intensity statin.²⁶ Similarly, a large US registry (GOULD, 2019–2021) found that after 2 years of usual care, merely ~14–15% of ASCVD patients attained LDL-C <1.4 mmol/l.²⁷ These findings are consistent with previously reported Indonesian data summarised in the background, which also show low attainment of guideline-recommended LDL-C targets in high-risk patients.

Our findings thus add to a growing body of evidence that the vast majority of real-world ASCVD patients are not achieving contemporary LDL-C goals, even in varied healthcare settings. This treatment gap is especially pertinent in low- and middle-income countries and Asian populations, where resource constraints and practice patterns may further limit aggressive lipid management.

Several factors may explain the difficulty in achieving LDL-C control in our cohort. Clinical inertia and gaps between guidelines and practice are likely contributors: not all eligible patients were prescribed high-intensity statins, and few received combination therapy. Some clinicians may also have targeted older, less aggressive LDL goals or been cautious with high-dose statins due to safety concerns. Historical data suggest Asian patients have sometimes been managed with lower statin doses due to heightened sensitivity or safety concerns.^21,28 While Asian individuals may achieve substantial LDL reductions at moderate doses, the prevalence of treatment ‘underdosing’ (or lack of therapy escalation) in our study suggests room for improvement in aligning with guideline-directed therapy. Encouragingly, our analysis demonstrates that using high-intensity statins can significantly improve the odds of goal attainment – a finding that supports more widespread adoption of intensive regimens barring contraindications.

Additionally, patient-related factors, such as adherence, play a critical role. We could not verify medication adherence in this study, but real-world adherence to statins is notoriously suboptimal. A pan-Asian survey reported that 44.1% of patients occasionally forgot to take their statin, and others discontinued due to perceived side-effects.¹¹ Such non-adherence may have masked the true efficacy of therapy, leading to falsely low goal attainment. Many ‘treated’ patients likely did not take medication consistently, highlighting adherence as a key target for intervention – through education, follow-up or fixed-dose combinations – to improve LDL-C control without new drugs.

Interestingly, hypertension was associated with lower odds of LDL-C goal attainment, possibly reflecting a more complex risk profile and frequent comorbidities. Competing treatment priorities or clinician caution with statin uptitration in the context of polypharmacy may also contribute. Supporting this notion, a study in Spain noted that patients with established CVD (and likely multiple risk factors) were significantly more likely to fail LDL targets despite intensive therapy.¹⁹ Similarly, a registry in northeast China found lower LDL goal attainment in patients with diabetes and those with prior CVD, among other factors.²¹ These trends indicate that patients at highest risk – those who would benefit most from aggressive LDL lowering – are often the least likely to reach targets, likely due to therapeutic inertia, complexity of care or higher baseline LDL-C. This highlights the importance of integrated management, ensuring that attention to hypertension does not detract from lipid control. Multidisciplinary secondary prevention programs, such as cardiac rehabilitation clinics, may help close these gaps by simultaneously addressing lifestyle, blood pressure, lipids and adherence.

From a public health perspective, these findings are relevant for Indonesia and comparable Asian healthcare settings. Indonesia is the world’s fourth most populous country and is experiencing a growing burden of non-communicable diseases. Within this burden, elevated LDL-C is a major modifiable contributor to ASCVD morbidity and mortality.²⁹ Yet, our study reveals that even among patients who have entered the healthcare system (in private hospitals) and have recognised ASCVD, optimal therapy is often not achieved. The situation is likely even more dire in the broader population. For example, a community-based study in Indonesia found that among adults at high cardiovascular risk, only ~1% were on statin therapy, reflecting huge treatment gaps in primary prevention.³⁰

Even in secondary prevention, before the era of PCSK9 inhibitors, registries showed low usage of high-intensity statins and adjuncts in many Asian countries.¹³ Our real-world data provide contemporary evidence that these therapy gaps persist, and they highlight a critical unmet need for improving guideline adherence and intensification of lipid-lowering therapy. If 88% of ASCVD patients are above the LDL-C goal (as in our study), that suggests a large proportion are likely experiencing unnecessary preventable risk for recurrent events. Bridging this gap could have substantial benefits; modelling studies and clinical trials have shown that achieving lower LDL-C levels (e.g. with intensive statins or combination therapy) yields further risk reductions without major safety trade-offs.^6,20

It is also important to consider therapeutic options beyond conventional statins. Guidelines recommend adding ezetimibe as the first intensification step, and PCSK9 inhibitors for those who still do not reach targets.²⁴ PCSK9 inhibitors, while highly efficacious in lowering LDL-C by an additional ~50–60%, are expensive and not widely accessible in Indonesia at present. Recent analyses suggest that most very-high-risk patients will require a PCSK9 inhibitor to achieve an LDL <1.4 mmol/l even after maximised statin–ezetimibe therapy.³¹ This aligns with our observation that high-intensity statins alone were often insufficient. However, without insurance coverage or price reductions, PCSK9 inhibitors remain out of reach for most patients. Although we were not able to perform a cost analysis, the economic barrier is clearly a major consideration. Our findings highlight the need for more affordable and scalable strategies in secondary prevention, such as ensuring patients receive maximally tolerated statin doses, expanding access to generic ezetimibe and advocating for inclusion of PCSK9 inhibitors in national formularies for the highest-risk groups.

Our study contributes valuable evidence from an understudied population and, to our knowledge, represents one of the first real-world analyses of LDL-C management in Indonesian private hospitals. The application of IPTW adjustment strengthens the comparative findings on statin intensity. Nonetheless, several limitations should be acknowledged.

First, the cross-sectional design precludes any inference of causality, and associations between factors, such as statin intensity or hypertension and LDL-C goal attainment, may be influenced by residual confounding. Although IPTW improved balance in measured covariates, unmeasured factors, such as dietary habits, physician treatment preferences or duration of prior statin therapy, may still bias results. Furthermore, LDL-C attainment was assessed using single available measurements without longitudinal follow-up in this cross-sectional study, preventing evaluation of trajectories over time or their impact on clinical outcomes. We did not perform inpatient-versus-outpatient subgroup analyses, because LDL-C measurement timing and statin initiation/intensification relative to hospitalisation were not standardised in the EMR extract, and such comparisons could be confounded by acuity and indication. The study’s retrospective nature may also give rise to potential selection bias and confounding.

Second, patient adherence to statins could not be verified; pharmacy refill-based adherence metrics (e.g. PDC/MPR) were not available in our dataset. Some individuals may have been misclassified as treated despite poor compliance, lowering observed goal attainment. Use of non-statin lipid-lowering agents, such as ezetimibe, was not systematically recorded and may have contributed to target achievement in a small subset.

Finally, the study was conducted in two private hospitals in Jakarta, which are urban and relatively resource-rich settings; therefore, the findings may not generalise to rural populations or those treated in the public healthcare system, where LDL-C control may be even poorer. Direct comparisons with the public healthcare system were not possible because our EMR extraction was limited to private hospitals and we did not have access to harmonised individual-level public-sector data for matched analyses. These limitations temper the conclusions, but the overall message is consistent: a substantial gap persists between guideline-recommended LDL-C targets and those achieved in routine practice. By highlighting this gap, our study underscores the need for intensified lipid-lowering therapy and improved care delivery in secondary prevention across Indonesia.

Conclusion

In two private hospitals in Jakarta, only 12.1% of ASCVD patients achieved LDL-C ≤1.4 mmol/l at index assessment, despite 66.1 and 33.7% of them being prescribed with moderate- and high-intensity statins, respectively. High-intensity statin therapy was associated with approximately twofold higher likelihood of target attainment versus moderate-intensity statin therapy after covariate adjustment, whereas most other clinical characteristics were not independently associated. These results highlight a substantial care gap in routine practice and inform the suboptimal outcomes of current lipid-lowering management. The study results also support the need for consistent implementation of guideline-directed lipid-lowering therapy, which includes improving clinician awareness for indication of high-intensity statin, and timely escalation beyond statin monotherapy when required. Future work should evaluate pragmatic strategies to improve treatment escalation, and formally assess the economic and implementation implications of treatment escalation within Indonesian private-hospital settings.

Click here to view Supplementary Material.