COVID-19 assessment in family practice-A clinical decision rule based on self-rated symptoms and contact history.

The study aimed to evaluate the diagnostic accuracy of contact history and clinical symptoms and to develop decision rules for ruling-in and ruling-out SARS-CoV-2 infection in family practice. We performed a prospective diagnostic study. Consecutive inclusion of patients coming for COVID-PCR testing to 19 general practices. Contact history and self-reported symptoms served as index test. PCR testing of nasopharyngeal swabs served as reference standard. Complete data were available from 1141 patients, 605 (53.0%) female, average age 42.2 years, 182 (16.0%) COVID-PCR positive. Multivariable logistic regression showed highest odds ratios (ORs) for "contact with infected person" (OR 9.22, 95% CI 5.61-15.41), anosmia/ageusia (8.79, 4.89-15.95), fever (4.25, 2.56-7.09), and "sudden disease onset" (2.52, 1.55-4.14). Patients with "contact with infected person" or "anosmia/ageusia" with or without self-reported "fever" had a high probability of COVID infection up to 84.8%. Negative response to the four items "contact with infected person, anosmia/ageusia, fever, sudden disease onset" showed a negative predictive value (NPV) of 0.98 (95% CI 0.96-0.99). This was present in 446 (39.1%) patients. NPV of "completely asymptomatic," "no contact," "no risk area" was 1.0 (0.96-1.0). This was present in 84 (7.4%) patients. To conclude, the combination of four key items allowed exclusion of SARS-CoV-2 infection with high certainty. With the goal of 100% exclusion of SARS-CoV-2 infection to prevent the spread of SARS-CoV-2 to the population level, COVID-PCR testing could be saved only for patients with negative response in all items. The decision rule might also help for ruling-in SARS-CoV-2 infection in terms of rapid assessment of infection risk.

Introduction

Rapid and accurate assessment of patients with suspected severe acute respiratory syndrome coronavirus type 2 (SARS-CoV-2) infection upon presentation to the primary care practices is of paramount importance to ensure optimal diagnostic and therapeutic management. Infected patients must be reliably identified in order to initiate quarantine measures and organize appropriate patient care. In addition, it is important not to misclassify uninfected individuals as infected, thereby frightening them and excluding them from work and social life. Efficient testing strategies are valuable not only to ensure accurate patient assessment but also to prevent the spread of SARS-CoV-2 at the population level. Therefore, a low-threshold testing strategy has been implemented in many countries, also in Germany[1,2]. Since early summer 2020, patients in Germany were allowed to receive a polymerase chain reaction (PCR) nasopharyngeal swab in their family physicians’ office if they had symptoms, felt uncertain about a possible infection, or needed a test result for legal reasons. Thus, access to testing was relatively easy, without selection of patients. This testing strategy has resulted in a large volume of COVID-PCR, which is time-consuming and costly.

Medical history and clinical signs and symptoms might contribute to an efficient selection of patients, in particular when decision rules can be developed to facilitate diagnostic decision making regarding test ordering of corona virus disease 2019 (COVID)-PCR. A systematic review including all studies from primary care and hospital settings has shown that most clinical signs and symptoms have high specificity but low sensitivity, thus making it difficult to exclude the disease[3]. Only four studies have been conducted in the primary care setting[4-7]. Anosmia[4-7], fever[5], and first grade contact with an infected person[4] was found to be predictive for SARS-CoV-2 infection. However, a combination of signs and symptoms were not evaluated in these studies, while the urgent need for prospective studies in an unselected population presenting to primary care has been identified[3]. The aim of the present study was to evaluate the diagnostic accuracy of anamnesis, contact history and clinical symptoms and to develop medical decision rules for ruling-in and ruling-out of SARS-CoV-2 infection in family practice.

Methods

Study design

This prospective diagnostic study was conducted from November 25, 2020 to February 26, 2021 in nineteen family practices in urban and rural areas of Upper Bavaria, Germany. The practices are part of a network that includes a total of 210 teaching practices at the Institute which are average practices representative for all practices in that area. All participating practices provide unselected primary care in the community for people making an initial approach to a medical professional within the social health insurance system. All patients (at least 18 years old) who came for COVID-PCR were asked to complete a short questionnaire on medical history, self-reported symptoms (SRS) and possible contact with (potentially) infected persons. Patients were included consecutively. The questionnaire items were generated from literature research and from information about the core symptoms of the disease as outlined by the website of the Robert Koch Institute[8]. The Robert Koch Institute (RKI) is the government’s central scientific institution in the field of biomedicine. Anamnestic items, contact history and SRS served as an index test (for queried symptoms, see Table 1). Nasopharyngeal swabs were performed and sent to the local medical laboratory for PCR analysis (reference standard). In Germany, the diagnostic decision is based on absolute detection of viral RNA in PCR analysis using a cycle threshold (Ct) of at least 40 cycles. The study was approved by the Ethics Committee of the Medical Faculty of the Technical University of Munich. All patients received written study information, and informed consent was obtained from all participants.

Table 1

Multivariable regression model of the questionnaire items.

Questionnaire item	Multivariable regression OR	p value
Demographics
(1) Sex = male	1.17 (0.79, 1.73)	0.436
(2) Age (in years)	1.03 (1.01, 1.05)	<0.001
Symptoms
(3) Anosmia/ageusia	8.79 (4.89, 15.95)	<0.001
(4) Fever—yes	4.25 (2.56, 7.09)	<0.001
5) Sudden disease onset	2.52 (1.55, 4.14)	<0.001
(6) Limp pain	1.72 (1.02, 2.91)	0.041
(7) Dry cough	1.69 (1.08, 2.62)	0.020
(8) Headache	1.14 (0.70, 1.84)	0.598
(9) Common cold	1.05 (0.67, 1.65)	0.822
(10) Fatigue	1.01 (0.58, 1.75)	0.978
(11) Diarrhea	0.91 (0.49, 1.61)	0.742
(12) Sore throat	0.52 (0.32, 0.83)	0.006
(13) Dyspnea	0.32 (0.14, 0.69)	0.005
Medical history
(14) Nicotin use	0.45 (0.25, 0.76)	0.004
(15) Chronic disease	0.34 (0.20, 0.57)	<0.001
Contact history
(16) Contact with infected person	9.22 (5.61, 15.41)	<0.001
(17) Stay in corona risk area	1.48 (0.67, 3.08)	0.310
(18) Contact with persons with suspected infection	1.28 (0.78, 2.07)	0.324

Variables in subcategories are ordered according to odds ratios (OR) of multivariable regression analysis (n = 1141). Intercept of the logistic regression model: −4.653

Statistical analysis

The distribution of continuous data is described by means and standard deviations. Qualitative data are presented by absolute and relative frequencies. Descriptive data were analyzed with t test or Chi-Square-test. Sensitivities, specificities, positive predictive values (PPV), negative predictive values (NPV), positive likelihood ratios, negative likelihood ratios, diagnostic odds ratios (ORs), and respective 95% confidence intervals (CIs) were computed for the items of the questionnaire. In some countries, high-risk contacts are dealt with in a separate testing strategy than symptomatic patients. Therefore, sensitivity analyses were performed without the “contact history” variables.

Investigated statistical models and machine learning methods were a conditional inference decision tree, a respective random forest, a Lasso model with the best cross-validated performance, a sparser Lasso model that did not perform significantly worse (using the 1se rule), and two multivariable logistic regression models built with and without Akaike Information Criterion-based stepwise backward variable selection[9]. The performance of diagnostic modeling was measured by the area under the receiver operating characteristics curve (AUC). A benchmark study was conducted to compare and internally validate the models’ performance by fivefold cross-validation. Thus, each model was repeatedly built on parts of the data (i.e., training data) and applied to independent parts of the data (i.e., test data) for an unbiased internal performance evaluation. For effect estimation and interpretation purposes, another conditional inference decision tree and a multivariable logistic regression model were refit to the whole data. ORs with 95% CIs are presented for the latter. Only patients with complete data were analyzed. For sample size calculation, we applied the rule of thumb 1:10 for the ratio of the model parameters to the number of observations in the less frequent outcome class[10]. Therefore, limiting sample size of 180 test positives had been determined a priori to allow consistent effect estimation of 18 model parameters in a multivariable logistic regression model[10]. More advanced rules, e.g., according to Riley et al.[11], involving the anticipated outcome proportion and model performance have not been applied due to the dynamic development of the pandemic, which did not allow reliable assumptions to be made a priori[11]. Significance of group differences and regression coefficients were assessed at exploratory two-sided alpha levels of 5%. Computations were conducted with R 4.0.3 (The R Foundation for Statistical Computing, Vienna, Austria).

As the PCR results were available to us on a daily basis, we were able to enroll patients into the study until the predefined sample size was reached. The data were entered twice by K.R. A comparison was made by inspection; in the event of a mismatch of variables, the information in the original questionnaire was checked and adopted in the data set. The statistical analysis was performed by the statistician A.H.

Table 2

Diagnostic measures of the questionnaire items and selected decision rules.

	Sensitivity	Specificity	Predictive values		Likelihood ratio		Diagnostic OR
			Positive	Negative	Positive	Negative
Demographics
(1) Sex = male	0.48 (0.41, 0.56)	0.53 (0.50, 0.56)	0.16 (0.13, 0.20)	0.84 (0.81, 0.87)	1.04 (0.88, 1.22)	0.97 (0.83, 1.13)	1.07 (0.78, 1.47)
(2) Age >39 years	0.60 (0.52, 0.67)	0.51 (0.48, 0.54)	0.19 (0.16, 0.22)	0.87 (0.84, 0.90)	1.22 (1.07, 1.40)	0.79 (0.65, 0.95)	1.56 (1.13, 2.15)
Symptoms
(3) Anosmia/ageusia	0.25 (0.19, 0.32)	0.94 (0.92, 0.96)	0.45 (0.35, 0.55)	0.87 (0.85, 0.89)	4.33 (3.03, 6.18)	0.79 (0.73, 0.86)	5.45 (3.55, 8.38)
(4) Fever—yes	0.34 (0.27, 0.41)	0.87 (0.84, 0.89)	0.33 (0.26, 0.40)	0.87 (0.85, 0.89)	2.55 (1.97, 3.31)	0.76 (0.68, 0.85)	3.35 (2.34, 4.80)
(5) Sudden disease onset	0.46 (0.39, 0.54)	0.71 (0.68, 0.74)	0.23 (0.19, 0.28)	0.87 (0.85, 0.90)	1.59 (1.32, 1.92)	0.76 (0.66, 0.87)	2.10 (1.52, 2.90)
(6) Limp pain	0.46 (0.39, 0.54)	0.72 (0.69, 0.75)	0.24 (0.20, 0.29)	0.88 (0.85, 0.90)	1.66 (1.38, 2.01)	0.75 (0.65, 0.86)	2.23 (1.62, 3.09)
(7) Dry cough	0.43 (0.36, 0.51)	0.69 (0.65, 0.71)	0.21 (0.17, 0.25)	0.86 (0.84, 0.89)	1.38 (1.14, 1.67)	0.83 (0.72, 0.94)	1.67 (1.21, 2.31)
(8) Headache	0.48 (0.41, 0.56)	0.61 (0.58, 0.64)	0.19 (0.15, 0.23)	0.86 (0.83, 0.89)	1.23 (1.04, 1.46)	0.85 (0.73, 0.99)	1.45 (1.05, 1.99)
(9) Common cold	0.43 (0.36, 0.50)	0.62 (0.59, 0.65)	0.18 (0.14, 0.22)	0.85 (0.82, 0.88)	1.14 (0.94, 1.37)	0.92 (0.80, 1.05)	1.24 (0.90, 1.71)
(10) Fatigue	0.54 (0.46, 0.61)	0.56 (0.53, 0.59)	0.19 (0.16, 0.23)	0.87 (0.84, 0.89)	1.23 (1.06, 1.44)	0.82 (0.69, 0.97)	1.50 (1.09, 2.07)
(11) Diarrhea	0.14 (0.10, 0.20)	0.85 (0.83, 0.87)	0.15 (0.10, 0.22)	0.84 (0.82, 0.86)	0.96 (0.66, 1.42)	1.01 (0.94, 1.07)	0.96 (0.61, 1.51)
(12) Sore throat	0.31 (0.25, 0.39)	0.60 (0.57, 0.63)	0.13 (0.10, 0.16)	0.82 (0.79, 0.85)	0.79 (0.63, 0.99)	1.14 (1.02, 1.28)	0.69 (0.49, 0.97)
(13) Dyspnea	0.07 (0.04, 0.12)	0.91 (0.88, 0.92)	0.12 (0.07, 0.20)	0.84 (0.81, 0.86)	0.75 (0.43, 1.32)	1.03 (0.98, 1.07)	0.73 (0.40, 1.34)
Medical history
(14) Nicotin use	0.15 (0.10, 0.21)	0.73 (0.70, 0.76)	0.09 (0.06, 0.13)	0.82 (0.79, 0.84)	0.55 (0.38, 0.79)	1.17 (1.08, 1.25)	0.47 (0.31, 0.73)
(15) Chronic disease	0.17 (0.12, 0.23)	0.72 (0.69, 0.75)	0.10 (0.07, 0.14)	0.82 (0.79, 0.85)	0.61 (0.44, 0.86)	1.15 (1.06, 1.24)	0.53 (0.35, 0.80)
Contact history
(16) Contact with infected person	0.58 (0.51, 0.65)	0.80 (0.77, 0.82)	0.35 (0.30, 0.41)	0.91 (0.89, 0.93)	2.86 (2.40, 3.41)	0.52 (0.44, 0.62)	5.46 (3.91, 7.63)
(17) Stay in corona risk area	0.08 (0.05, 0.13)	0.92 (0.90, 0.94)	0.17 (0.10, 0.26)	0.84 (0.82, 0.86)	1.07 (0.63, 1.82)	0.99 (0.95, 1.04)	1.07 (0.60, 1.92)
(18) Contact with persons with suspected infection	0.38 (0.31, 0.46)	0.83 (0.81, 0.86)	0.30 (0.25, 0.37)	0.88 (0.85, 0.90)	2.31 (1.83, 2.91)	0.74 (0.66, 0.83)	3.12 (2.21, 4.40)
Combinations
At least one symptom or condition #3–18	1.00 (0.98, 1.00)	0.09 (0.07, 0.10)	0.17 (0.15, 0.20)	1.00 (0.96, 1.00)	1.10 (1.07, 1.12)	0.00 (0.00, -)	–
At least one positive response to “contact with infected person”, anosmia/ageusia, fever, sudden disease onset	0.96 (0.92, 0.98)	0.46 (0.42, 0.49)	0.25 (0.22, 0.28)	0.98 (0.96, 0.99)	1.76 (1.65, 1.88)	0.10 (0.05, 0.19)	18.29 (8.90, 37.57)
Combinations excluding “contact with infected person” (sensitivity analysis)
At least one symptom or condition #17	0.91 (0.86, 0.95)	0.16 (0.14, 0.19)	0.17 (0.15, 0.20)	0.91 (0.85, 0.95)	1.09 (1.03, 1.15)	0.54 (0.33, 0.88)	2.03 (1.18, 3.49)
At least one positive response to anosmia/ageusia, fever, sudden disease onset	0.70 (0.63, 0.77)	0.62 (0.59, 0.66)	0.26 (0.22, 0.30)	0.92 (0.89, 0.94)	1.87 (1.65, 2.12)	0.48 (0.38, 0.60)	3.94 (2.80, 5.56)