Amber Fozia Khan
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
Amber Fozia Khan has a total of 9 co-authors. Here are the top co-authors by number of shared publications:
- Kiran Saqib
School of Public Health Sciences, University of Waterloo, Waterloo, Canada
2 shared publications - Zahid Ahmad Butt
University of Waterloo
2 shared publications Anne O'RIordan
1 shared publication- Jawairia Mohammed1 shared publication
- Jennifer Shuldiner
Research and Innovation Institute, Women’s College Hospital, Toronto, Canada
1 shared publication Nadine Hare
1 shared publication
Latest Publications
- Open Access
- 1 June 2026
- Published by JMIR Publications Inc. in JMIR Preprints
Designing Compassion into Asynchronous Messaging in Primary Care: Findings from Co-Creation Workshops (Preprint)
BACKGROUND Asynchronous messaging (AM), including patient messages, lab results, specialist correspondence, and internal team communications, is now routine in primary care, offering patients timely access while reducing the need for in-person visits. However, AM has generated substantial workload and is thought to contribute to clinician burnout. Despite its increasing role in primary care, limited empirical work has examined how AM can be intentionally designed to sustainably support compassionate relational care. OBJECTIVE This study aimed to co-design compassionate and sustainable approaches to asynchronous electronic messaging in primary care by engaging patients, clinicians, clinic staff, and health system leaders in identifying challenges and developing feasible solutions for communication workflows, standards, and platform design. METHODS We conducted sequential co-creation workshops between June 2025 and January 2026 as part of a multi-phase, design-led qualitative study. Separate workshops were held with (i) patients; (ii) medical office assistants, nurses, and staff; and (iii) family physicians; followed by (iv) a final multi-stakeholder workshop including patients, clinicians, health system leaders, and digital health stakeholders. Workshops built on prior ethnographic observations and semi-structured interviews across multiple primary care clinics. Structured ideation methods and digital collaboration tools were used to generate, cluster, and refine ideas, which were iteratively synthesized. RESULTS Participants (N=69) from each group described a shared need for convenient, timely communication outside clinical encounters, but focused on different opportunities and challenges. Patients (n) emphasized access, usability, and frustration with opaque messaging processes. Clinic staff (n) highlighted message volume, triage complexity, and workflow burden, while family physicians (n) emphasized emotional labour, tension between patient-centred care and burnout, and the invisibility of messaging work. Through the co-creation process, participants identified three interrelated solution areas: (1) an information-flow management role or software to triage, categorize, and prioritize messages; (2) electronic communication standards that clarify appropriate use, response expectations, and responsibilities; and (3) user-centred messaging platforms . In the final multi-stakeholder workshop, participants underscored that sustainable funding, provincial policy direction, and coordinated governance are prerequisites for implementation. CONCLUSIONS AM can support compassionate primary care when messaging is guided by shared standards, supported by active inbox management, and embedded in team-based, well-designed platforms. Clear roles, expectations, and enabling infrastructure may reduce clinician hesitation, expand capacity, and strengthen equitable, patient-centered care. CLINICALTRIAL Not applicable
- 1 June 2026
- Published by JMIR Publications Inc. in JMIR Preprints
Open AccessDesigning Compassion into Asynchronous Messaging in Primary Care: Findings from Co-Creation Workshops (Preprint)
BACKGROUND Asynchronous messaging (AM), including patient messages, lab results, specialist correspondence, and internal team communications, is now routine in primary care, offering patients timely access while reducing the need for in-person visits. However, AM has generated substantial workload and is thought to contribute to clinician burnout. Despite its increasing role in primary care, limited empirical work has examined how AM can be intentionally designed to sustainably support compassionate relational care. OBJECTIVE This study aimed to co-design compassionate and sustainable approaches to asynchronous electronic messaging in primary care by engaging patients, clinicians, clinic staff, and health system leaders in identifying challenges and developing feasible solutions for communication workflows, standards, and platform design. METHODS We conducted sequential co-creation workshops between June 2025 and January 2026 as part of a multi-phase, design-led qualitative study. Separate workshops were held with (i) patients; (ii) medical office assistants, nurses, and staff; and (iii) family physicians; followed by (iv) a final multi-stakeholder workshop including patients, clinicians, health system leaders, and digital health stakeholders. Workshops built on prior ethnographic observations and semi-structured interviews across multiple primary care clinics. Structured ideation methods and digital collaboration tools were used to generate, cluster, and refine ideas, which were iteratively synthesized. RESULTS Participants (N=69) from each group described a shared need for convenient, timely communication outside clinical encounters, but focused on different opportunities and challenges. Patients (n) emphasized access, usability, and frustration with opaque messaging processes. Clinic staff (n) highlighted message volume, triage complexity, and workflow burden, while family physicians (n) emphasized emotional labour, tension between patient-centred care and burnout, and the invisibility of messaging work. Through the co-creation process, participants identified three interrelated solution areas: (1) an information-flow management role or software to triage, categorize, and prioritize messages; (2) electronic communication standards that clarify appropriate use, response expectations, and responsibilities; and (3) user-centred messaging platforms . In the final multi-stakeholder workshop, participants underscored that sustainable funding, provincial policy direction, and coordinated governance are prerequisites for implementation. CONCLUSIONS AM can support compassionate primary care when messaging is guided by shared standards, supported by active inbox management, and embedded in team-based, well-designed platforms. Clear roles, expectations, and enabling infrastructure may reduce clinician hesitation, expand capacity, and strengthen equitable, patient-centered care. CLINICALTRIAL Not applicable
- Open Access
- 24 November 2021
- Published by JMIR Publications Inc. in JMIR Mental Health
- Vol. 8 (11) , e29838
Machine Learning Methods for Predicting Postpartum Depression: Scoping Review
Background: Machine learning (ML) offers vigorous statistical and probabilistic techniques that can successfully predict certain clinical conditions using large volumes of data. A review of ML and big data research analytics in maternal depression is pertinent and timely, given the rapid technological developments in recent years. Objective: This study aims to synthesize the literature on ML and big data analytics for maternal mental health, particularly the prediction of postpartum depression (PPD). Methods: We used a scoping review methodology using the Arksey and O’Malley framework to rapidly map research activity in ML for predicting PPD. Two independent researchers searched PsycINFO, PubMed, IEEE Xplore, and the ACM Digital Library in September 2020 to identify relevant publications in the past 12 years. Data were extracted from the articles’ ML model, data type, and study results. Results: A total of 14 studies were identified. All studies reported the use of supervised learning techniques to predict PPD. Support vector machine and random forest were the most commonly used algorithms in addition to Naive Bayes, regression, artificial neural network, decision trees, and XGBoost (Extreme Gradient Boosting). There was considerable heterogeneity in the best-performing ML algorithm across the selected studies. The area under the receiver operating characteristic curve values reported for different algorithms were support vector machine (range 0.78-0.86), random forest method (0.88), XGBoost (0.80), and logistic regression (0.93). Conclusions: ML algorithms can analyze larger data sets and perform more advanced computations, which can significantly improve the detection of PPD at an early stage. Further clinical research collaborations are required to fine-tune ML algorithms for prediction and treatment. ML might become part of evidence-based practice in addition to clinical knowledge and existing research evidence.
- 24 November 2021
- Published by JMIR Publications Inc. in JMIR Mental Health
- Vol. 8 (11) , e29838
Open AccessMachine Learning Methods for Predicting Postpartum Depression: Scoping Review
Background: Machine learning (ML) offers vigorous statistical and probabilistic techniques that can successfully predict certain clinical conditions using large volumes of data. A review of ML and big data research analytics in maternal depression is pertinent and timely, given the rapid technological developments in recent years. Objective: This study aims to synthesize the literature on ML and big data analytics for maternal mental health, particularly the prediction of postpartum depression (PPD). Methods: We used a scoping review methodology using the Arksey and O’Malley framework to rapidly map research activity in ML for predicting PPD. Two independent researchers searched PsycINFO, PubMed, IEEE Xplore, and the ACM Digital Library in September 2020 to identify relevant publications in the past 12 years. Data were extracted from the articles’ ML model, data type, and study results. Results: A total of 14 studies were identified. All studies reported the use of supervised learning techniques to predict PPD. Support vector machine and random forest were the most commonly used algorithms in addition to Naive Bayes, regression, artificial neural network, decision trees, and XGBoost (Extreme Gradient Boosting). There was considerable heterogeneity in the best-performing ML algorithm across the selected studies. The area under the receiver operating characteristic curve values reported for different algorithms were support vector machine (range 0.78-0.86), random forest method (0.88), XGBoost (0.80), and logistic regression (0.93). Conclusions: ML algorithms can analyze larger data sets and perform more advanced computations, which can significantly improve the detection of PPD at an early stage. Further clinical research collaborations are required to fine-tune ML algorithms for prediction and treatment. ML might become part of evidence-based practice in addition to clinical knowledge and existing research evidence.
- Open Access
- 22 April 2021
- Published by JMIR Publications Inc. in JMIR Preprints
Machine Learning Methods for Predicting Postpartum Depression: Scoping Review (Preprint)
BACKGROUND Machine learning (ML) offers vigorous statistical and probabilistic techniques that can successfully predict certain clinical conditions using large volumes of data. A review of ML and big data research analytics in maternal depression is pertinent and timely, given the rapid technological developments in recent years. OBJECTIVE This study aims to synthesize the literature on ML and big data analytics for maternal mental health, particularly the prediction of postpartum depression (PPD). METHODS We used a scoping review methodology using the Arksey and O’Malley framework to rapidly map research activity in ML for predicting PPD. Two independent researchers searched PsycINFO, PubMed, IEEE Xplore, and the ACM Digital Library in September 2020 to identify relevant publications in the past 12 years. Data were extracted from the articles’ ML model, data type, and study results. RESULTS A total of 14 studies were identified. All studies reported the use of supervised learning techniques to predict PPD. Support vector machine and random forest were the most commonly used algorithms in addition to Naive Bayes, regression, artificial neural network, decision trees, and XGBoost (Extreme Gradient Boosting). There was considerable heterogeneity in the best-performing ML algorithm across the selected studies. The area under the receiver operating characteristic curve values reported for different algorithms were support vector machine (range 0.78-0.86), random forest method (0.88), XGBoost (0.80), and logistic regression (0.93). CONCLUSIONS ML algorithms can analyze larger data sets and perform more advanced computations, which can significantly improve the detection of PPD at an early stage. Further clinical research collaborations are required to fine-tune ML algorithms for prediction and treatment. ML might become part of evidence-based practice in addition to clinical knowledge and existing research evidence.
- 22 April 2021
- Published by JMIR Publications Inc. in JMIR Preprints
Open AccessMachine Learning Methods for Predicting Postpartum Depression: Scoping Review (Preprint)
BACKGROUND Machine learning (ML) offers vigorous statistical and probabilistic techniques that can successfully predict certain clinical conditions using large volumes of data. A review of ML and big data research analytics in maternal depression is pertinent and timely, given the rapid technological developments in recent years. OBJECTIVE This study aims to synthesize the literature on ML and big data analytics for maternal mental health, particularly the prediction of postpartum depression (PPD). METHODS We used a scoping review methodology using the Arksey and O’Malley framework to rapidly map research activity in ML for predicting PPD. Two independent researchers searched PsycINFO, PubMed, IEEE Xplore, and the ACM Digital Library in September 2020 to identify relevant publications in the past 12 years. Data were extracted from the articles’ ML model, data type, and study results. RESULTS A total of 14 studies were identified. All studies reported the use of supervised learning techniques to predict PPD. Support vector machine and random forest were the most commonly used algorithms in addition to Naive Bayes, regression, artificial neural network, decision trees, and XGBoost (Extreme Gradient Boosting). There was considerable heterogeneity in the best-performing ML algorithm across the selected studies. The area under the receiver operating characteristic curve values reported for different algorithms were support vector machine (range 0.78-0.86), random forest method (0.88), XGBoost (0.80), and logistic regression (0.93). CONCLUSIONS ML algorithms can analyze larger data sets and perform more advanced computations, which can significantly improve the detection of PPD at an early stage. Further clinical research collaborations are required to fine-tune ML algorithms for prediction and treatment. ML might become part of evidence-based practice in addition to clinical knowledge and existing research evidence.