Risk factors and nomogram-based prediction of dry eye disease following intraocular lens implantation in cataract patients

Abstract

Background

Postoperative dry eye disease (DED) is a common complication following cataract surgery with intraocular lens (IOL) implantation, adversely affecting ocular comfort and visual outcomes. Identifying risk factors and developing predictive models may facilitate targeted interventions and improved patient management.

Methods

A retrospective analysis was conducted involving 376 cataract patients who underwent phacoemulsification with IOL implantation between January 2021 and December 2024. Patients were divided into a DED group (n = 158) and a non- DED group (n = 218) based on clinical evaluation, including the Ocular Surface Disease Index (OSDI), tear meniscus height (TMH), tear film break-up time (BUT), Schirmer I test (SIt), and fluorescein staining (FL) scores. Risk factors were identified using univariate and multivariate logistic regression analyses. A nomogram was constructed from significant predictors and assessed by receiver operating characteristic (ROC) analysis, bootstrap internal validation, and decision curve analysis (DCA).

Results

The incidence of postoperative DED was 42.02% (Among the DED patients, 66.46% had mild DED, and 33.54% had moderate DED). Univariate analysis revealed that patients with DED were significantly older and had a higher prevalence of orthokeratology lens wear, diabetes, and rheumatoid arthritis, along with worse nuclear hardness grading, prolonged surgical time, abnormal ocular surface stress test (OSST) results, and impaired ocular surface parameters (all p < 0.05). Multivariate analysis identified orthokeratology lens wear (OR 3.472, p = 0.018), diabetes (OR 3.193, p = 0.016), diminished meibomian gland secretion (OR 3.228, p = 0.033), increased meibum viscosity (OR 3.548, p = 0.018), and elevated conjunctivochalasis grade (OR 3.092, p = 0.027) as independent predictors. The nomogram demonstrated an AUC of 0.761, with a sensitivity of 71.67% and specificity of 82.18%. Internal validation yielded a corrected C-index of 0.785, and DCA confirmed its clinical utility.

Conclusions

In this study, orthokeratology lens wear, diabetes, and adverse ocular surface parameters, including diminished meibomian gland secretion, increased meibum viscosity, and elevated conjunctivochalasis grades were statistically associated with the occurrence of postoperative DED following IOL implantation. The nomogram developed from these associations demonstrated moderate discriminatory ability (AUC = 0.761) and good calibration. Prospective studies and external validation in diverse populations are warranted to confirm the model’s generalizability and predictive performance.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12886-025-04369-1.

Introduction

Dry eye disease (DED) following intraocular lens (IOL) implantation in cataract patients has become a prevalent postoperative complication, necessitating thorough investigation into its risk factors and predictive strategies. Cataract surgery, one of the most frequently performed ophthalmic procedures worldwide, has advanced significantly with the development of phacoemulsification and modern IOL technologies [1–3]. Despite these improvements, the onset of DED postoperatively remains a significant clinical concern, potentially compromising visual outcomes and patient satisfaction. The pathogenesis of postoperative DED is multifactorial [4, 5]. involving surgical trauma, disruption of corneal innervation, postoperative inflammation, and altered tear film dynamics. Additional contributors include preexisting ocular surface conditions, intraoperative parameters such as incision size and surgical duration, and postoperative management protocols. Moreover, systemic comorbidities—such as diabetes mellitus and autoimmune diseases—and environmental exposures (e.g., low humidity, prolonged screen use) may further predispose patients to DED following IOL implantation [6–8].

Risk stratification is essential for optimizing surgical outcomes and improving patient care. Traditional risk assessment approaches typically focus on isolated clinical parameters; however, integrating multiple risk factors into a comprehensive predictive model offers a more robust framework for anticipating postoperative complications. In this context, nomogram-based models have gained increasing attention, as they enable individualized risk prediction by incorporating diverse clinical variables. Derived from multivariate statistical analyses, nomograms provide a graphical tool that can be easily applied in routine clinical practice. These models not only enhance evidence-based decision-making but also facilitate the personalization of perioperative management strategies aimed at reducing adverse outcomes [9, 10].

This study aims to conduct a comprehensive analysis of risk factors associated with DED following IOL implantation in cataract patients and to develop a nomogram-based predictive model. By systematically evaluating preoperative, intraoperative, and postoperative variables, we seek to identify significant predictors of DED and quantify their relative contributions. The integration of these factors into a nomogram is anticipated to enhance risk stratification and support the implementation of tailored management protocols. Ultimately, this research endeavors to contribute to the optimization of cataract surgery outcomes by reducing the incidence and severity of postoperative DED, thereby advancing patient care within the field of ophthalmology.

Methods

Study design

This retrospective study evaluated patients diagnosed with cataract who underwent intraocular lens implantation at our institution between January 2021 and December 2024. A total of 376 patients were included and subsequently stratified into two cohorts based on the postoperative development of DED: 158 patients who developed DED and 218 who did not. The study design, methodology, and analytical protocols were developed in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines [11]. Informed consent was obtained from all participants. The study was reviewed and approved by the ethics committee of Tianjin Eye Hospital and conducted in accordance with the Declaration of Helsinki and relevant institutional guidelines. All participant data were anonymized prior to analysis to ensure confidentiality.

Inclusion and exclusion criteria

Inclusion criteria for this study were: (1) patients aged 18 years or older with a clinical diagnosis of cataract; (2) patients who underwent phacoemulsification combined with intraocular lens implantation; (3) availability of comprehensive preoperative and postoperative clinical data, including ocular surface evaluations; and (4) a minimum postoperative follow-up duration of three months. Exclusion criteria comprised: (1) a history of prior ocular surgery; (2) preexisting DED or other ocular surface disorders; (3) concurrent ocular pathologies, such as glaucoma, uveitis, or corneal dystrophies; (4) systemic conditions known to affect tear production or ocular surface integrity (e.g., Sjögren’s syndrome, rheumatoid arthritis); and (5) incomplete clinical records or insufficient follow-up data.

Surgical technique

All procedures were performed by a dedicated ophthalmic team at our institution. Following a comprehensive preoperative evaluation, patients were placed in the supine position and administered local anesthesia. A small incision was made via a limbal tunnel, and sodium hyaluronate was injected into the anterior chamber as a viscoelastic agent. A continuous curvilinear capsulorhexis was subsequently performed, followed by ultrasound phacoemulsification, nuclear fragmentation, and aspiration of both the opaque lens nucleus and cortical material. The anterior chamber was then irrigated thoroughly, after which additional viscoelastic was injected prior to the implantation of an appropriately powered intraocular lens. The IOL was carefully positioned, and the anterior chamber was re-irrigated to remove any residual debris, with the incision sealed using a watertight closure technique. Postoperative management included standard bandaging and administration of anti-infective therapy. Follow-up examinations were scheduled at 7 days, 1 month, and 3 months postoperatively, with the assessment of DED conducted at the 3-month follow-up.

Data collection and outcome measures

Patient data were comprehensively collected, including demographic characteristics (gender and age), cataract duration, history of orthokeratology lens wear, smoking and alcohol consumption, and comorbid conditions (hypertension, diabetes mellitus, dyslipidemia, and rheumatoid arthritis). Additionally, surgical and ocular parameters were recorded: lens nucleus hardness graded according to the Emery-Little standard, patient disease awareness, surgical duration, incision length and number, intraoperative cumulative dissipated energy (CDE), and postoperative ocular surface stress test (OSST) results. Postoperative evaluations at 7 days included best corrected visual acuity (BCVA), measured using a NIDEK ARK-1 auto-refractometer with small-pupil correction using trial lenses, and intraocular pressure assessment. Further assessments encompassed postoperative psychological status—evaluated using the Hospital Anxiety and Depression Scale (HADS)—as well as medication adherence and comprehensive ocular surface evaluations [12]. The meibomian gland secretion score was determined based on a grading system adapted from Shimazaki et al. [13], in which digital pressure was applied to the upper and lower tarsal plates, and secretion quantity was scored as follows: 0 when secretions were observed in all five glands, 1 when present in three to four glands, 2 when present in one to two glands, and 3 when no secretions were observed.

Diagnostic criteria for DED

Comprehensive evaluation of both subjective symptoms and objective ocular surface assessments, based on the Chinese Expert Consensus on DED: Definition and Classification (2020). Subjective evaluation included patient-reported discomfort, such as sensations of dryness, foreign body presence, and burning, which were deemed mandatory for diagnosis. In addition, the Ocular Surface Disease Index (OSDI) questionnaire was administered, consisting of seven items assessing ocular discomfort over the preceding week, with a total score ranging from 0 to 100; an OSDI score of 25 or greater was considered indicative of DED. Objective assessments comprised the measurement of tear meniscus height, tear film break-up time (BUT), Schirmer I test (SIt), and corneal fluorescein staining (FL) score. The tear meniscus height was determined under a slit-lamp examination following the instillation of 2% fluorescein sodium into the conjunctival sac. For the BUT evaluation, after instilling 2% fluorescein sodium and allowing the patient to blink 3–4 times, the interval until the first disruption in the tear film was recorded, with a value of less than 10 s serving as a critical diagnostic criterion. The SIt was performed by placing a folded filter paper strip at the lateral one-third of the lower conjunctival fornix, with the patient keeping their eyes closed for 5 min; a wetting length of less than 5 mm was considered abnormal. Corneal fluorescein staining was graded immediately after the BUT test using a slit-lamp: a score of 0 indicated no staining, 1 denoted punctate staining, 2 represented diffuse punctate staining, and 3 corresponded to patchy staining, with any positive FL score supporting the diagnosis.

Statistical analysis

Statistical analyses were conducted using SPSS 26.0 and R 3.6.3. Continuous variables were expressed as mean ± standard deviation (x̄ ± SD) and compared between groups using independent-samples t-tests. Categorical variables were presented as counts (percentages) and analyzed using χ² tests, continuity-corrected χ² tests, or Fisher’s exact test as appropriate. To mitigate the influence of multicollinearity, intercorrelations among predictive factors were assessed and adjusted. Multiple logistic regression analysis was performed to identify risk factors for DED, with odds ratios (ORs) and 95% confidence intervals (CIs) calculated. The rms package was utilized to develop a nomogram and generate receiver operating characteristic (ROC) curves, while the consistency index (C-index) evaluated model discrimination. The optimal cutoff was determined via the Youden index, and calibration was assessed using the Hosmer–Lemeshow test along with bootstrap resampling (n = 1,000) to derive a corrected C-index and calibration plots. Decision curve analysis (DCA) further evaluated the clinical utility of the model. A P-value < 0.05 was considered statistically significant.

Results

Postoperative DED incidence and clinical characteristics

At 3 months post-discharge, 376 patients who underwent cataract surgery with IOL implantation were evaluated. Among them, 158 patients (42.02%) developed postoperative DED. Patients in the DED group reported symptoms including dryness, foreign body sensation, and burning, which were corroborated by abnormal ocular surface assessments. The OSDI scores were significantly elevated, whereas objective parameters—tear meniscus height (TMH), tear film BUT, and SIt values—were markedly reduced compared to the non-DED group (P < 0.05). In contrast, fluorescein staining (FI) scores were significantly higher in the DED group (P < 0.05) (Table 1). Within the DED cohort, disease severity was further stratified based on symptomatology and clinical findings. Mild DED was observed in 105 patients (66.46%), whereas 53 patients (33.54%) presented with moderate DED. Following targeted health education and symptomatic management, most patients demonstrated either complete resolution or notable improvement in symptoms.

Table 1.

Comparison of ocular surface parameters between dry eye and non-dry eye groups

Parameter	Dry Eye Group (n = 158)	Non-Dry Eye Group (n = 218)	t-value	p-value
TMH (mm)	0.17 ± 0.05	0.20 ± 0.06	5.126	< 0.001
BUT (s)	4.46 ± 1.25	10.82 ± 2.39	30.55	< 0.001
SIt (mm/5min)	3.43 ± 1.04	8.20 ± 1.66	31.86	< 0.001
FI Score (points)	2.75 ± 0.68	1.11 ± 0.38	29.78	< 0.001
OSDI (points)	35.10 ± 6.65	29.10 ± 4.35	10.57	< 0.001

TMH Tear meniscus height, BUT Tear film break-up time, SIt Schirmer I test, FI Fluorescein staining score, OSDI, Ocular Surface Disease Index

Univariate analysis of factors associated with postoperative DED

Table 2 summarizes the univariate analysis comparing demographic, clinical, and surgical variables between patients with and without postoperative DED following IOL implantation. No statistically significant differences were found between the two groups in terms of gender, cataract duration, smoking and drinking history, hypertension, hyperlipidemia, disease awareness, CDE, incision length, number of incisions, best-corrected visual acuity (BCVA), postoperative intraocular pressure (IOP), postoperative anxiety/depression, or medication compliance (all p > 0.05). In contrast, significant differences were observed for several variables. Patients in the DED group were more likely to be aged ≥ 60 years (p = 0.002) and had higher frequencies of orthokeratology lens wear (p < 0.001), diabetes (p < 0.001), and rheumatoid arthritis (p < 0.001). Nuclear hardness grade, surgical time, and OSST outcomes also differed significantly between groups (all p < 0.001). Additionally, ocular surface parameters—including meibomian gland secretion score, meibum viscosity grade, and conjunctivochalasis grade—were significantly worse among patients who developed DED (all p < 0.001) (Table 2).

Table 2.

Univariate analysis of factors associated with postoperative dry eye in cataract patients undergoing IOL implantation

Factor	Dry Eye Group (n = 158)	Non-Dry Eye Group (n = 218)	χ²/t Value	p-Value
Gender (Male/Female)	75 (47.47%)/83 (52.53%)	114 (52.29%)/104 (47.71%)	2.51	0.113
Cataract Duration (months)	7.15 ± 2.00	6.80 ± 1.90	1.15	0.252
Smoking History (Yes)	30 (18.99%)	32 (14.68%)	2.92	0.085
Drinking History (Yes)	33 (20.89%)	36 (16.51%)	3.23	0.071
Hypertension (Yes)	47 (29.75%)	60 (27.52%)	0.56	0.456
Hyperlipidemia (Yes)	50 (31.65%)	60 (27.52%)	3.15	0.075
Disease Awareness (Low/Medium–High)	74 (46.84%)/84 (53.16%)	108 (49.54%)/110 (50.46%)	0.73	0.394
CDE	4.20 ± 1.30	4.05 ± 1.20	1.63	0.132
Incision Length (mm)	2.50 ± 0.54	2.45 ± 0.50	0.95	0.352
Number of Incisions (1/2)	95 (60.13%)/63 (39.87%)	135 (61.93%)/83 (38.07%)	1.05	0.305
BCVA	0.65 ± 0.19	0.63 ± 0.18	1.62	0.132
Postoperative IOP (mmHg)	17.50 ± 3.10	17.70 ± 3.20	0.79	0.421
Postoperative Anxiety/Depression (Yes)	39 (24.68%)	57 (26.15%)	0.55	0.486
Postoperative Medication Compliance	Poor 24 (15.19%)/Good 134 (84.81%)	Poor 29 (13.30%)/Good 189 (86.70%)	1.31	0.313
Age (≥ 60/<60)	115 (63.29%)/43 (36.71%)	125 (57.34%)/93 (42.66%)	9.47	0.002
Orthokeratology Lens Wear (Yes)	55 (34.81%)	15 (6.88%)	53.82	< 0.001
Diabetes (Yes)	84 (53.16%)	46 (21.10%)	36.08	< 0.001
Rheumatoid Arthritis (Yes)	38 (24.05%)	30 (13.76%)	18.93	< 0.001
Nuclear Hardness (I–II/III–VI)	82 (51.90%)/76 (48.10%)	87 (39.91%)/131 (60.09%)	11.35	0.001
Surgical Time (min)	22.30 ± 5.60	18.80 ± 4.70	10.63	< 0.001
OSST (Negative/Positive)	131 (82.91%)/27 (17.09%)	201 (92.20%)/17 (7.80%)	18.24	< 0.001
Meibomian Gland Secretion Score (0–1/2–3)	58 (36.71%)/100 (63.29%)	159 (73.00%)/59 (27.00%)	14.53	< 0.001
Meibum Viscosity Grade (1–2/3–5)	31 (19.62%)/127 (80.38%)	117 (53.67%)/101 (46.33%)	139.55	< 0.001
Conjunctivochalasis Grade (1–2/3)	59 (37.34%)/99 (62.66%)	164 (75.23%)/54 (24.77%)	141.43	< 0.001

OSST, Ocular surface stress test, BCVA, Best corrected visual acuity, IOP Intraocular pressure, CDE, Cumulative dissipated energy

Multivariate analysis of risk factors for postoperative DED

Multivariate logistic regression analysis identified several independent predictors of postoperative DED. Significant risk factors included orthokeratology lens wear (OR 3.472, 95% CI 1.265–9.590, p = 0.018) and diabetes (OR 3.193, 95% CI 1.290–7.920, p = 0.016). In addition, ocular surface parameters were strongly associated with DED, with the meibomian gland secretion score (OR 3.228, 95% CI 1.115–9.165, p = 0.033), meibum viscosity grade (OR 3.548, 95% CI 1.285–9.970, p = 0.018), and conjunctivochalasis grade (OR 3.092, 95% CI 1.145–8.050, p = 0.027) all significantly increasing the risk. Although factors such as age, rheumatoid arthritis, nuclear hardness grade, surgical time, and OSST showed trends toward higher risk, they did not reach statistical significance (Table 3).

Table 3.

Multivariate logistic regression analysis of risk factors for dry eye

Factors	β Value	Standard Error Value	Wald Value	OR Value	95% CI for OR	p-Value
Age	0.917	0.547	2.96	2.515	0.906–6.995	0.095
Rheumatoid Arthritis	1.103	0.56	3.4	3.004	0.955–9.445	0.069
Nuclear Hardness Grade	1.38	0.76	2.99	3.945	0.848–18.290	0.089
Surgical Time	1.208	0.665	2.91	3.342	0.868–12.790	0.093
OSST	1.265	0.702	2.9	3.545	0.858–14.685	0.091
Orthokeratology Lens Wear	1.255	0.536	5.75	3.472	1.265–9.590	0.018
Diabetes	1.156	0.479	6.14	3.193	1.290–7.920	0.016
Meibomian Gland Secretion Score	1.173	0.556	4.56	3.228	1.115–9.165	0.033
Meibum Viscosity Grade	1.278	0.545	5.83	3.548	1.285–9.970	0.018
Conjunctivochalasis Grade	1.129	0.515	4.91	3.092	1.145–8.050	0.027

OSST Ocular surface stress test, β Regression coefficient, OR Odds ratio, CI Confidence interval

Nomogram development and validation for predicting postoperative DED

A predictive nomogram for postoperative DED following IOL implantation was constructed based on five independent risk factors identified by multivariate logistic regression. For each variable, points were allocated according to its value on the nomogram’s scale. The total score, calculated as the sum of these individual points, corresponds to the estimated probability of developing DED; a higher cumulative score indicates an elevated risk (Fig. 1).

Fig. 1.

Nomogram for predicting dry eye disease (DED) in cataract patients following intraocular lens (IOL) implantation. This nomogram integrates multiple risk factors, including orthokeratology lens wear, diabetes, meibomian gland secretion score, meibum viscosity grade, and conjunctivochalasis grade. For each factor, the corresponding point value is assigned. The total score, derived from summing the individual points, predicts the likelihood of developing DED postoperatively. The predicted probability of DED is displayed on the lower axis. A higher total score correlates with an increased risk of developing DED after IOL implantation. DED: Dry eye disease; IOL: Intraocular lens

The discriminatory performance of the nomogram was evaluated by generating a ROC curve, yielding an AUC of 0.761 (95% CI, 0.657–0.835). At the optimal threshold determined by the maximum Youden index, the model demonstrated a sensitivity of 71.67% and a specificity of 82.18%, reflecting its moderate ability to distinguish between patients who do and do not develop DED postoperatively (Fig. 2).

Fig. 2.

Receiver operating characteristic (ROC) curve illustrating the performance of the nomogram in predicting dry eye disease (DED) in cataract patients after IOL implantation. The ROC curve evaluates the discriminative ability of the nomogram, with the area under the curve (AUC) indicating the model’s accuracy. The nomogram demonstrates an AUC of 0.761 (95% CI: 0.657–0.835), reflecting its moderate ability to distinguish between patients who develop DED and those who do not following intraocular lens (IOL) implantation. The reference line (dashed) represents a random model with an AUC of 0.5. A higher AUC indicates better model performance

Internal validation was performed using bootstrap resampling (n = 1,000 iterations) to assess the model’s calibration. The corrected C-index was 0.785 (95% CI, 0.735–0.831), indicating a good level of concordance between predicted and observed outcomes. Furthermore, the Hosmer-Lemeshow goodness-of-fit test produced a χ² value of 2.651 (p = 0.873), corroborating the model’s calibration accuracy. Calibration curves further supported that the predicted probabilities closely aligned with the actual incidence of DED (Fig. 3).

Fig. 3.

Calibration curve demonstrating the agreement between predicted and observed probabilities of dry eye disease (DED) in cataract patients post-IOL implantation. The calibration curve compares the predicted probabilities of DED, as calculated by the nomogram, with the actual observed outcomes in the study cohort. The ideal curve (dashed pink line) represents perfect calibration, where predicted probabilities perfectly match the observed probabilities. The apparent curve (solid green line) shows the uncorrected observed probabilities, while the bias-corrected curve (yellow line) adjusts for overfitting and provides a more accurate reflection of model performance. A closer alignment of the bias-corrected curve with the ideal curve indicates better model calibration

The clinical utility of the nomogram was assessed via DCA. In the DCA, the horizontal line represents the scenario where no patients develop DED and thus receive no intervention (net benefit = 0), while the oblique line reflects the outcome if all patients were assumed to develop DED and were treated accordingly, resulting in a negative net benefit. The net benefit derived from the nomogram was substantially higher than these extreme scenarios across a range of threshold probabilities, underscoring its potential clinical applicability in guiding postoperative management (Fig. 4).

Fig. 4.

Decision curve analysis (DCA) evaluating the clinical net benefit of the nomogram for predicting dry eye disease (DED) in cataract patients following IOL implantation. The x-axis represents threshold probabilities, and the y-axis indicates net clinical benefit. The green curve (“Nomogram Prediction Model”) shows the net benefit across a range of thresholds, compared with the blue line (“All”), assuming all patients develop DED, and the pink line (“None”), assuming none do. The nomogram demonstrated higher net benefit than either treat-all or treat-none strategies across relevant threshold probabilities, supporting its clinical utility

Discussion

DED is a common complication following cataract surgery with IOL implantation, substantially impacting postoperative comfort and visual outcomes. Our study confirms that postoperative DED is a multifactorial condition, influenced by both patient-specific characteristics and intraoperative factors. In univariate analysis, several demographic and clinical variables—such as advanced age, history of orthokeratology lens wear, diabetes, and rheumatoid arthritis—were significantly associated with DED development. Additionally, indicators of ocular surface integrity, including nuclear hardness grade, surgical duration, OSST outcomes, meibomian gland secretion score, meibum viscosity grade, and conjunctivochalasis grade, were notably impaired in patients who developed DED. These findings highlight the complex interplay between systemic comorbidities, ocular surface health, and surgical parameters in the pathogenesis of postoperative DED. In multivariate logistic regression, orthokeratology lens wear and diabetes were identified as independent risk factors, while ocular surface parameters—namely, meibomian gland secretion score, meibum viscosity grade, and conjunctivochalasis grade—remained strong predictors. Although variables such as age, rheumatoid arthritis, nuclear hardness, surgical time, and OSST exhibited trends toward increased risk, they did not reach statistical significance in the multivariate model. The final nomogram, developed based on these key predictors, demonstrated moderate discriminative ability (AUC = 0.761) and good calibration (corrected C-index = 0.785), supporting its potential utility in clinical practice for individualized risk prediction and targeted perioperative management.

The significant association between orthokeratology lens wear and postoperative DED may be partially attributed to the mechanical and environmental stress exerted by contact lens use on the ocular surface. Prolonged orthokeratology wear can induce chronic microtrauma to the corneal epithelium, destabilize the tear film, and trigger inflammatory changes within the ocular surface microenvironment. These effects may impair tear film quality and compromise meibomian gland function, thereby increasing vulnerability to DED following cataract surgery. Additionally, cessation of lens wear during the perioperative period may further disrupt tear film homeostasis, compounding the observed risk [10, 14]. Diabetes, another independent risk factor identified in our study, is known to affect the ocular surface through multiple mechanisms. Diabetic microangiopathy and neuropathy can impair lacrimal gland function, resulting in aqueous tear deficiency. Moreover, systemic inflammation and oxidative stress associated with diabetes may exacerbate ocular surface inflammation and disrupt tear film stability. Altered lipid metabolism in diabetic patients can also impair meibomian gland function, leading to changes in meibum composition and increased viscosity, which further degrade tear film integrity [15, 16]. Collectively, these pathophysiological mechanisms contribute to the elevated risk of postoperative DED in diabetic patients, consistent with our findings.

The ocular surface parameters evaluated in this study—including meibomian gland secretion score, meibum viscosity grade, and conjunctivochalasis grade—serve as direct indicators of ocular surface integrity and function. Dysfunction of the meibomian glands is a primary contributor to evaporative DED, as these glands secrete the lipid layer essential for reducing tear evaporation [17]. Our findings demonstrate that patients with impaired meibomian gland secretion and elevated meibum viscosity grades exhibited significantly higher odds of developing DED. Likewise, increased conjunctivochalasis—defined by redundant conjunctival folds—can disrupt the uniform distribution of the tear film and impair tear drainage, thereby promoting tear film instability [18, 19]. The strong associations between these parameters and postoperative DED highlight the critical role of comprehensive preoperative ocular surface evaluation in patients undergoing IOL implantation. Interestingly, although advanced age and rheumatoid arthritis showed trends toward increased DED risk, they did not reach statistical significance in the multivariate model. The lack of significance for age may stem from its indirect effects on ocular surface health, which are more accurately captured by specific parameters such as meibomian gland function [20]. Similarly, the impact of rheumatoid arthritis—an autoimmune condition that can promote ocular surface inflammation—may have been attenuated by confounding factors in the multivariate analysis. These results suggest that while such systemic conditions may contribute to DED susceptibility, their effects are likely mediated through more direct ocular surface dysfunctions [21, 22].

The development of a nomogram based on the identified independent risk factors offers a practical tool for clinicians. By integrating demographic and ocular surface parameters into a single predictive model, the nomogram allows for the individualized estimation of postoperative dry eye risk. The model’s moderate discrimination and robust calibration indicate that it can reliably stratify patients according to their risk profiles, thereby facilitating targeted preoperative counseling and the implementation of preventive strategies. The decision curve analysis further supports the clinical utility of the nomogram, as it demonstrated a net benefit across a range of threshold probabilities. This suggests that incorporating the nomogram into clinical practice could improve postoperative outcomes by enabling early intervention in high-risk patients (for further details, please refer to Supplementary Materials S1).

Koh [23]and Chaudhary et al.‘s [24] studies suggest that contact lens wear is associated with dry eye symptoms and discomfort. Both acknowledge that prolonged wear can exacerbate symptoms in some individuals, but neither study definitively claims that contact lens use always induces DED. In contrast, Ruan et al.‘s [25] research suggests that Ortho-K lenses have minimal impact on the ocular surface, particularly regarding tear film stability and meibomian gland function, indicating that these lenses may not directly induce DED. However, Xie et al.‘s [26]study highlights that Ortho-K lenses can lead to ocular discomfort and transient dry eye symptoms, particularly in the early stages of wear. While these symptoms may resolve over time, this study underscores the potential for dry eye-related discomfort during the initial wear period, which contradicts the idea that Ortho-K lenses do not cause dry eye symptoms. In our study, orthokeratology lens wear was considered a potential risk factor for postoperative DED due to its known association with ocular surface changes that may increase the likelihood of developing DED after cataract surgery. Therefore, including this factor as a predictor for postoperative DED aligns with the well-established link between contact lens use and DED. However, as a retrospective study, our findings should be interpreted with caution. The observed association between Ortho-K lens wear and postoperative DED may reflect indirect effects, and causality cannot be established from this analysis. This association should not be regarded as definitive, and further prospective studies and external validation are necessary to confirm these results and refine our understanding of the underlying mechanisms.

This study has several limitations that warrant consideration. First, its retrospective and single-center design may introduce selection bias and restrict causal inference, thereby limiting the generalizability of the findings to broader populations. Although multiple risk factors were identified and incorporated into a nomogram, unmeasured confounding variables—such as environmental exposures, lifestyle habits, and genetic predispositions—were not accounted for and may have influenced the observed outcomes. Moreover, some ocular surface assessments relied on semi-quantitative or subjective evaluations, which may introduce interobserver variability. The absence of external validation and the geographically homogeneous study population further limit the nomogram’s applicability to other settings or ethnic groups. Future multi-center, prospective studies with larger and more diverse cohorts are essential to validate and refine the model and to assess its clinical utility across different populations and healthcare environments. In addition, due to sample size constraints, we did not stratify patients into finer age categories or model dry eye severity as a continuous outcome. Further studies should explore age-specific risk stratification and the predictive value of disease severity. While our current analysis distinguished between mild and moderate dry eye, future research should incorporate the severe category with longer follow-up to capture the full disease spectrum and its progression. Finally, the implementation of objective, standardized diagnostic tools and evaluation of tailored interventions based on individual risk profiles may enhance both predictive accuracy and clinical management of postoperative DED.

Conclusions

In this retrospective analysis, orthokeratology lens wear, diabetes, and adverse ocular surface parameters, including diminished meibomian gland secretion, increased meibum viscosity, and elevated conjunctivochalasis grades were statistically associated with the occurrence of postoperative DED following IOL implantation. The nomogram developed from these associations demonstrated moderate discriminatory ability (AUC = 0.761) and good calibration. Prospective studies and external validation in diverse populations are warranted to confirm the model’s generalizability and predictive performance.

Authors’ contributions

Hou-Yin Xu, Sheng-Kun Lang, and Hao Jiang were responsible for the conceptualization, data curation, formal analysis, methodology, resource acquisition, and software implementation of the study. Hou-Yin Xu drafted the original manuscript and also performed the review and editing.

Funding

This study was supported by the Tianjin Key Medical Discipline Construction (Grant No. TJYXZDXK-3-004A-3).

Data availability

The datasets used and/or analyzed during the present study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

This study was approved by the Ethics Committee of Tianjin Eye Hospital. All procedures adhered to relevant guidelines and regulations, in accordance with the Declaration of Helsinki. Participant data were anonymized prior to analysis to ensure confidentiality and privacy. Informed consent was obtained from all subjects.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.