Impact of China’s diagnosis-intervention packet payment reform on pediatric pneumonia hospitalization costs: an interrupted time series analysis

Abstract

Background

Pediatric pneumonia remains a major cause of morbidity and mortality, imposing substantial financial burdens on healthcare systems and families. This study evaluates the impact of China’s diagnosis-intervention packet (DIP) payment reform on hospitalization costs and care quality for pediatric pneumonia.

Methods

We retrospectively analyzed hospitalization cost data from a pilot hospital for DIP reform, between January 2019 and December 2023. Cases were categorized into re-reform and post-reform phases based on DIP implementation. Interrupted time series regressions assessed immediate and long-term cost trends and clinical outcomes.

Results

A total of 13,133 pediatric pneumonia hospitalizations were included (4,053 pre-reform; 9,080 post-reform). Median hospitalization costs decreased from 4,150.7 RMB to 3,853.3 RMB, with the most notable reductions in medication costs (261.1 RMB) and comprehensive service fees (103.9 RMB). Interrupted time series analysis showed significant immediate reductions in comprehensive service costs (23.2%, P < 0.001) and medication costs (15.8%, P = 0.031), followed by sustained monthly declines in all types of hospitalization costs. Concurrently, clinical outcomes improved: cure rates increased significantly from 87.0 to 90.6% (P < 0.001) without increased ICU transfers (3.5% vs. 4.6%, P = 0.478).

Conclusion

The implementation of DIP payment reform was associated with significant reductions in hospitalization costs for pediatric pneumonia while maintaining key quality indicators such as cure rates and ICU transfer frequencies. The observed cost reductions were primarily driven by lower medication expenses and reduced use of unnecessary diagnostic services, reflecting a shift towards value-based care. These findings underscore the potential of DIP reform to enhance hospital efficiency and financial sustainability without compromising patient care.

Introduction

Pediatric pneumonia remains a leading cause of morbidity and mortality in children worldwide, placing a significant economic burden on both healthcare systems and families [1, 2]. In China, approximately 21 million children are diagnosed with pneumonia annually, resulting in 74,000 deaths [3]. Previous studies have highlighted the rising healthcare costs associated with pediatric pneumonia, as treatment often requires prolonged hospitalization, diagnostics, and medications [4]. These costs contribute significantly to both public healthcare expenditures and the financial strain on families.

In response to escalating costs, various reforms have been introduced to control medical costs, including diagnosis-related group payments and bundled payment models. In November 2020, the National Healthcare Security Administration launched a nationwide pilot program for diagnosis-intervention packet (DIP) payment reform, with the goal of improving healthcare quality and controlling medical costs. The DIP model combines diagnosis-specific bundled payments with a total budget control mechanism. Under this system, healthcare providers receive a fixed payment for a defined package of services related to a specific diagnosis, such as pediatric pneumonia. The reform is designed to incentivize hospitals to deliver care more efficiently by encouraging cost-effective treatments while maintaining or improving clinical outcomes. It aims to control rising medical costs by reducing unnecessary procedures and hospitalizations, and aligning financial incentives with quality of care. This hybrid model is particularly relevant for low- and middle-income countries with fragmented health systems [5]. However, there is limited evidence on the long-term effectiveness of these reforms in pediatric care settings. Most existing studies focus on macrolevel data and aggregate outcomes [6, 7], but they lack a detailed analysis of cost components and their specific impact on patient groups, such as those with pediatric pneumonia.

This study evaluated the impact of DIP payment reform on both overall and detailed hospitalization costs for pediatric pneumonia patients. By assessing the effectiveness of the reform, we aim to provide valuable insights that can help hospitals refine their management strategies and improve cost-efficiency, particularly in the context of rising healthcare costs.

Methods

Study setting and DIP payment reform

This study was conducted at Shiyan Maternal and Child Health Hospital, a large, publicly funded institution specializing in pediatric and maternal care that serves over one million residents. This hospital is a pilot hospital for DIP payment reform in the region, underscoring its pivotal role in addressing the challenges of rising healthcare costs while maintaining high-quality care. The DIP integrates a total budget cap with a point-based payment system driven by big data and diagnosis-related grouping [8]. It emphasizes value-based purchasing and rational resource allocation through standardized disease grouping directories, disease-specific scoring, and dynamic adjustment of settlement weight coefficients for medical institutions. Although it shares operational similarities with the international DRG system, DIP’s simpler technological and management requirements render it particularly well suited for the current healthcare environment in China.

Data

We collected medical records of pediatric pneumonia inpatients from the hospital’s information system between 2019 and 2023. The inclusion criteria were as follows: (1) primary diagnosis of pneumonia (ICD-10 codes J09-J18, P23) with matching admission and discharge diagnoses and (2) age under 14 years. The exclusion criteria were as follows: (1) comorbidities such as congenital heart defects, myocardial damage, and hand, foot, and mouth disease; (2) hospital length of stay ≤ 24 h; and (3) incomplete or incorrect medical record information.

Baseline information of the inpatients was collected, including sex, age, admission date, discharge date, primary and secondary diagnoses, corresponding ICD-10 codes for each diagnosis, and detailed hospitalization costs. Hospitalization costs were categorized into five types: (1) comprehensive medical service costs, including general medical service costs, general treatment operation costs, nursing costs, etc.; (2) diagnostic costs, including pathology diagnostic costs, laboratory diagnostic costs, imaging diagnostic costs, etc.; (3) medication costs, including Western medicine costs, traditional Chinese medicine costs, herbal medicine costs; (4) material costs, including one-time medical materials costs for examinations, and one-time medical material costs for treatment, one-time medical material costs for surgery, etc.; and (5) other costs, including treatment costs, blood fees and blood product costs, and other uncategorized costs.

Cost adjustment

On the basis of the implementation time of DIP payment reform in the pilot hospital (April 1, 2021), the data were divided into pre-reform (January 1, 2019, to March 31, 2021) and post-reform (April 1, 2021, to December 31, 2023) periods. To address inflation-driven cost variations, hospitalization expenditures were standardized via annual consumer price indices issued by the National Bureau of Statistics, with 2023 designated as the reference year (Supplementary Table 1). This inflation adjustment protocol eliminated temporal price fluctuations, enabling valid cross-year comparisons of cost control effects under the DIP payment system.

Statistical analysis

Categorical variables (e.g., demographic characteristics) were summarized as frequencies with percentages. Continuous variables (e.g., cost-related variables) were assessed for normality via the Kolmogorov-Smirnov tests. Comparisons between pre- and post-reform were performed using the Pearson’s 𝜒² tests for categorical variables and Wilcoxon rank-sum tests for continuous variables. All hypothesis tests were two-sided, with a significance level set at α = 0.05. R 4.3.2 software was used for data preprocessing and statistical analysis.

Interrupted time series regression (ITSR) models were established to assess the effect of DIP payment reform on both the level and trend of hospitalization costs for pediatric pneumonia inpatients. The ITSR model is a quasi-experimental research design used to evaluate the effects of interventions [9]. It measures the immediate and trend changes of interventions by collecting data before and after the intervention at multiple time points. This model has been widely used in healthcare research [10,11,12,13,14]. The expression of the model is as follows:

$$\:{Y}_{t}={\beta\:}_{0}+{\beta\:}_{1}\times\:time+{\beta\:}_{2}\times\:intervention+{\beta\:}_{3}\times\:post{+\epsilon\:}_{t}$$

In this study, the cost variables follow positively skewed distributions. Therefore, the outcome variable $\:{Y}_{t}$ represents the natural logarithm of hospitalization costs. The time variable spans 60 consecutive months labelled 1 to 60. The binary intervention variable indicates policy implementation status (0 = pre-reform; 1 = post-reform). The variable post denotes the time series after the reform. Pre-reform observations (n = 27) were coded as 0, whereas post-reform observations (n = 33) were sequentially labelled 1 to 33. $\:{\epsilon\:}_{t}$ denotes the random error. $\:{\beta\:}_{0}$ quantifies the baseline value, $\:{\beta\:}_{1}$ the pre-reform trend, $\:{\beta\:}_{2}$ the immediate reform effect, and $\:{\beta\:}_{3}$ the sustained effect. The post-reform trend is captured by $\:{{\beta\:}_{1}+\beta\:}_{3}$.

To address the impacts of COVID-19 and seasonal variations in pediatric pneumonia incidence, we introduced two key adjustments to the model. First, a binary indicator variable was defined (1 = January–December 2020; 0 = other periods) to capture systemic disruptions caused by COVID-19 containment measures [15]. Second, Fourier harmonic terms (sine/cosine pairs with annual periodicity) were incorporated into the model to remove the seasonal effects [9]. The Durbin-Watson test was applied to check for first-order autocorrelation in the error term [16]. By conducting significance tests on the regression coefficients, the statistical significance of changes in pediatric pneumonia hospitalization costs before and after DIP reform was assessed.

Results

Study population characteristics​​ and pre-post comparisons

The study included a total of 13,133 hospitalized patients, with 4,053 patients before the implementation of the DIP payment reform and 9,080 patients after. The baseline characteristics of the patients and their distributions before and after the DIP payment reform are displayed in Table 1. There was no statistically significant difference in sex distribution between patients before and after the DIP payment reform (P = 0.320). The highest proportion of patients were under one year old both before and after policy implementation, with the difference in age distribution being nonsignificant (P = 0.089).

Table 1 Comparison of baseline characteristics and hospitalization costs for pediatric pneumonia patients before and after the DIP payment reform

A substantial proportion of patients had hospital stays of less than one week. A statistically significant difference was observed in the length of hospital stay before and after the DIP payment reform (P < 0.001), with the percentage of patients staying longer than one week decreasing from 44.1% prior to the reform to 31.0% afterward. ICU transfer rates remained stable between pre- and post-DIP periods (3.5% vs. 4.6%, P = 0.478). Discharge outcomes significantly improved, with the cured rate increasing from 87.0 to 90.6% (P < 0.001), while rates of uncured cases (0.3% vs. 0.2%) and other outcomes (0.4% vs. 0.1%) showed no adverse trends. The antibiotic cost ratio (proportion of total drug costs attributed to antibiotics) decreased significantly from 0.60 (± 0.04) pre-DIP to 0.59 (± 0.05) post-DIP (P < 0.001).

Cost outcomes

Kolmogorov-Smirnov normality tests indicated that the cost-related variables followed nonnormal distributions (P < 0.001). Thus, they are summarized as medians with interquartile ranges (Table 1). The median hospitalization cost per case decreased significantly from 4,150.7 RMB pre-reform to 3,853.3 RMB post-reform. The cost components exhibited divergent trends: medication cost‌ demonstrated the most substantial reduction (median decrease: 261.1 RMB), followed by the comprehensive service cost (median decrease: 103.9 RMB)‌. In contrast, the diagnostic cost and material cost showed compensatory increases. The average medication cost ratio decreased from 27% before the reform to 23% after the reform (P < 0.001).

Figure 1 shows the monthly changes in the average hospitalization cost and individual expense categories, along with the fitted curves of the ISTR model. The dashed line denotes the ISTR model, whereas the solid line indicates the seasonally adjusted ISTR model. Table 2 presents the seasonally adjusted parameter estimates from the ISTR models for cost-related outcomes. Before the reform, the average monthly growth rates for diagnostic cost showed a significant upward trend, with an increase of 1.5% (95% CI: 0.7–2.4%; P < 0.001). Pre-reform data indicated a marginal decline in material cost, contrasted with incremental increases across other cost categories. However, none of these trends were statistically significant.

Fig. 1

Monthly changes in hospitalization costs per visit for pediatric pneumonia patients: (A) Total cost, (B) Comprehensive medical service cost, (C) Diagnostic cost, (D) Medication cost, (E) Material cost, and (F) Other cost

Table 2 Effects of DIP pay reform on hospitalization costs in pediatric pneumonia patients: interrupted time series log-linear regression model results

Immediately following the implementation of the DIP payment reform, the comprehensive medical service cost decreased by 23.2% (95% CI: 0.7–2.4%; P < 0.001), and the medication cost decreased by 15.8% (95% CI: 1.5–28.0%; P = 0.031). However, no statistically significant immediate changes were observed for diagnostic cost, material cost, or other costs.

After the reform, downward trends were observed in total hospitalization cost, diagnostic cost, medication cost, material cost, and other cost, with average monthly reductions of 0.7% (95% CI: 0.1–1.2%; P = 0.012), 1.0% (95% CI: 0.5–1.6%; P < 0.001), 0.8% (95% CI: 0.4–1.3%; P < 0.001), 1.8% (95% CI: 0.3–3.3%; P = 0.018), and 1.6% (95% CI: 0.4–2.8%; P = 0.010), respectively. Although there was an average monthly reduction of 0.3% in diagnostic cost, this trend was not statistically significant (P = 0.347).

Healthcare utilization trends

The ITSR analysis revealed distinct temporal patterns in care delivery following DIP implementation (Fig. 2; Table 3). Prior to the reform, the average hospital length of stay remained relatively stable (pre-reform trend slope: β = 0.02; 95% CI: -0.01 to 0.04; P = 0.176). Following the reform, a downward trend in length of stay was observed (post-reform trend slope: β = -0.03, 95% CI: -0.05 to 0; P = 0.062). The difference in trends between the pre- and post-reform periods was significant (P = 0.031). Concurrently, cure rates demonstrated accelerated improvement post-reform. While the baseline pre-DIP trend showed non-significant growth (β = 0.18, P = 0.129), the post-reform period showed a significant positive slope of 0.19 (95% CI: 0.05–0.34; P = 0.009). For cost metrics, medication cost ratios exhibited persistent yet non-significant declines both before (β = -0.16, P = 0.070) and after (β = -0.06, P = 0.283) the reform. Antibiotic cost ratios showed no significant baseline trend (β = 0.10, P = 0.563) or post-reform trajectory (β = -0.12, P = 0.234).

Fig. 2

Monthly changes in healthcare utilization for pediatric pneumonia patients: (A) average hospital length of stay, (B) cure rate, (C) medication cost ratio, and (D) antibiotic cost ratio

Table 3 Effects of DIP pay reform on length of stay and medication cost ratio in pediatric pneumonia patients: interrupted time series linear regression model results

Discussion

Our study shows that DIP payment reform is linked to significant reductions in hospitalization costs for pediatric pneumonia patients. The reform brought about both immediate and ongoing decreases in medication expenses and service fees, despite modest increases in diagnostic and material costs. Moreover, the observed shortening of hospital stays points to improvements in clinical efficiency. The observed cost reductions alongside stable ICU transfer rates suggest that DIP may balance efficiency and quality.

Our results indicate that comprehensive medical service costs, drug costs, and diagnostic costs represent a major share of the total hospitalization expenses for pediatric pneumonia patients. Notably, the medication cost ratio decreased from approximately 27% before the reform to 23% afterward—a finding that likely reflects the distinct treatment strategies employed in pediatric care. Owing to limitations in terms of age, cognitive level, and expressive ability, children often find it difficult to accurately describe their symptoms and disease conditions or communicate their posttreatment feelings and changes directly with healthcare providers [17]. Therefore, pediatric diseases rely more on diagnostic tests to strengthen clinical arguments, and children require more clinical care. In our study, the medication cost ratios before and after the reform were 27% and 23%, respectively, which are much lower than the drug proportion for adult pneumonia [18]. Due to the physiological characteristics and different responses to medications in children, lower doses of drugs are usually used to treat pneumonia. Additionally, compared with adult pneumonia, which requires more antibiotics, pediatric pneumonia may rely more on supportive or conservative treatments.

Following the DIP payment reform, we observed a significant downward trend in overall hospitalization costs. The interrupted time series analysis revealed an immediate reduction in comprehensive service and medication costs, along with a sustained decline in total costs over time. This suggests that the reform effectively curtailed the escalation of hospitalization expenses and helped alleviate the economic burden on families.

At the beginning of the reform, diagnostic and material costs did not show a significant immediate change. However, a clear downward trend emerged in subsequent months. This delayed effect may be attributed to the initial adaptation period required by hospitals to adjust to new financial incentives, as well as potential delays in government compensation [19]. At the beginning of DIP implementation, hospitals might not have rationalized the reform policies, and government financial compensation to public hospitals was not timely, causing short-term economic losses that impacted hospital operations [20]. Consequently, at the early stage of the DIP reform, hospitals might have increased diagnostic and material use to compensate for the economic losses caused by the reform. However, in the long term, compared with the period before DIP policy implementation, there was a significant downward trend in diagnostic costs and material costs, possibly related to the elimination of medical material markups and clinical pathway reform policies in China [21]. Therefore, despite the overall high levels of diagnostic and material costs, recent comprehensive reforms in the Chinese healthcare system have significantly impacted patient cost structures, leading to continuous and long-term optimization.

Interestingly, while diagnostic, drug, material, and other ancillary costs significantly decreased after the reform, comprehensive medical service costs—closely tied to total hospitalization expenses—did not exhibit a significant post-reform trend. This contrasts with findings from DRG reforms in adult populations, where reductions in procedural costs tend to dominate [22]. The preservation of service-related costs under the DIP model likely reflects its emphasis on valuing clinical labor, a critical consideration for healthcare systems in low- and middle-income countries facing workforce shortages. By ensuring that the intrinsic value of clinical treatment, consultation, and nursing is maintained, the DIP model supports the technical labor of healthcare personnel and encourages the adoption of high-risk, technically demanding procedures [23]. This means that the DIP model focuses more on the essence of medical services, the severity of the disease, and the complexity of treatment. Ensuring the price of medical services helps motivate healthcare personnel. To enhance competitiveness, hospitals need to establish a connection between medical services and performance, encouraging doctors to undertake new technologies and projects with high risk and technical difficulty to continuously improve the hospital’s medical level [24].

The proportion of pediatric pneumonia patients whose hospital stays exceeded one week significantly declined after the reform. Under a fixed health insurance budget, hospitals appear to have adopted strategies—such as reducing the length of stay and increasing bed turnover—to optimize resource utilization and maximize reimbursement under the new payment model. Crucially, ICU transfer rates remained stable (Δ = +1.1%, P = 0.478), while cure rates improved from 87.0 to 90.6% (P < 0.001). These dual outcomes suggest that DIP reforms achieved cost containment without compromising critical care utilization. Furthermore, the significant reduction in antibiotic cost ratios (P < 0.001) suggests optimized stewardship under DIP, potentially through reduced empiric broad-spectrum use.

Our study’s strengths include a long observation period and a large sample size, enabling robust control of historical trend changes. However, the single-center design limits the generalizability of the findings to other healthcare settings. Additionally, although factors such as sex, age, and hospital stay duration were considered, important confounding factors, such as disease severity and complications, were not fully accounted for. Third, this study did not assess post-discharge outcomes such as readmission rates or long-term complications, which may reflect unintended consequences of shortened hospital stays. Future studies should incorporate data from diverse hospitals and adjust for additional variables to increase the generalizability and accuracy of the findings.

Conclusion

The DIP payment reform underscores the potential of value-based payment systems to reduce economic burdens while maintaining, and possibly even enhancing, care quality in pediatric settings. Policymakers should ensure that cost control measures are complemented by timely financial support and robust monitoring systems to safeguard clinical quality during transitional periods. Future research should explore the broader applicability of the DIP model across diverse healthcare settings and patient populations, further examining how hospitals can balance cost containment with sustained improvements in care quality.